anhtld commited on
Commit
d78d4fe
·
verified ·
1 Parent(s): 877aa4c

Upload env_clip aggregate and dominance artifacts

Browse files
This view is limited to 50 files because it contains too many changes.   See raw diff
Files changed (50) hide show
  1. workspace/runs/ctt_base_context_obs_dominance_envclip_train_to_test/command.txt +1 -0
  2. workspace/runs/ctt_base_context_obs_dominance_envclip_train_to_test/config.yaml +11 -0
  3. workspace/runs/ctt_base_context_obs_dominance_envclip_train_to_test/data_hash.txt +6 -0
  4. workspace/runs/ctt_base_context_obs_dominance_envclip_train_to_test/eval.log +3 -0
  5. workspace/runs/ctt_base_context_obs_dominance_envclip_train_to_test/git_hash.txt +1 -0
  6. workspace/runs/ctt_base_context_obs_dominance_envclip_train_to_test/metrics.json +0 -0
  7. workspace/runs/ctt_base_context_obs_dominance_envclip_train_to_test/metrics_by_seed.json +0 -0
  8. workspace/runs/ctt_base_context_obs_dominance_envclip_train_to_test/metrics_by_task.json +157 -0
  9. workspace/runs/ctt_base_context_obs_dominance_envclip_train_to_test/report.md +16 -0
  10. workspace/runs/ctt_base_context_obs_dominance_envclip_train_to_test/split_hash.txt +14 -0
  11. workspace/runs/ctt_base_context_obs_dominance_envclip_train_to_test/table.tex +8 -0
  12. workspace/runs/ctt_base_context_obs_dominance_envclip_train_to_test/train.log +4 -0
  13. workspace/runs/ctt_base_context_obs_learned_dominance_basic_envclip_train_to_test/command.txt +1 -0
  14. workspace/runs/ctt_base_context_obs_learned_dominance_basic_envclip_train_to_test/config.yaml +11 -0
  15. workspace/runs/ctt_base_context_obs_learned_dominance_basic_envclip_train_to_test/data_hash.txt +6 -0
  16. workspace/runs/ctt_base_context_obs_learned_dominance_basic_envclip_train_to_test/eval.log +3 -0
  17. workspace/runs/ctt_base_context_obs_learned_dominance_basic_envclip_train_to_test/git_hash.txt +1 -0
  18. workspace/runs/ctt_base_context_obs_learned_dominance_basic_envclip_train_to_test/metrics.json +0 -0
  19. workspace/runs/ctt_base_context_obs_learned_dominance_basic_envclip_train_to_test/metrics_by_seed.json +2762 -0
  20. workspace/runs/ctt_base_context_obs_learned_dominance_basic_envclip_train_to_test/metrics_by_task.json +117 -0
  21. workspace/runs/ctt_base_context_obs_learned_dominance_basic_envclip_train_to_test/report.md +17 -0
  22. workspace/runs/ctt_base_context_obs_learned_dominance_basic_envclip_train_to_test/split_hash.txt +14 -0
  23. workspace/runs/ctt_base_context_obs_learned_dominance_basic_envclip_train_to_test/table.tex +8 -0
  24. workspace/runs/ctt_base_context_obs_learned_dominance_basic_envclip_train_to_test/train.log +4 -0
  25. workspace/runs/ctt_base_context_obs_learned_dominance_context_envclip_train_to_test/command.txt +1 -0
  26. workspace/runs/ctt_base_context_obs_learned_dominance_context_envclip_train_to_test/config.yaml +11 -0
  27. workspace/runs/ctt_base_context_obs_learned_dominance_context_envclip_train_to_test/data_hash.txt +6 -0
  28. workspace/runs/ctt_base_context_obs_learned_dominance_context_envclip_train_to_test/eval.log +3 -0
  29. workspace/runs/ctt_base_context_obs_learned_dominance_context_envclip_train_to_test/git_hash.txt +1 -0
  30. workspace/runs/ctt_base_context_obs_learned_dominance_context_envclip_train_to_test/metrics.json +0 -0
  31. workspace/runs/ctt_base_context_obs_learned_dominance_context_envclip_train_to_test/metrics_by_seed.json +2762 -0
  32. workspace/runs/ctt_base_context_obs_learned_dominance_context_envclip_train_to_test/metrics_by_task.json +117 -0
  33. workspace/runs/ctt_base_context_obs_learned_dominance_context_envclip_train_to_test/report.md +17 -0
  34. workspace/runs/ctt_base_context_obs_learned_dominance_context_envclip_train_to_test/split_hash.txt +14 -0
  35. workspace/runs/ctt_base_context_obs_learned_dominance_context_envclip_train_to_test/table.tex +8 -0
  36. workspace/runs/ctt_base_context_obs_learned_dominance_context_envclip_train_to_test/train.log +4 -0
  37. workspace/runs/ctt_base_context_obs_learned_dominance_context_success_envclip_train_to_test/command.txt +1 -0
  38. workspace/runs/ctt_base_context_obs_learned_dominance_context_success_envclip_train_to_test/config.yaml +11 -0
  39. workspace/runs/ctt_base_context_obs_learned_dominance_context_success_envclip_train_to_test/data_hash.txt +6 -0
  40. workspace/runs/ctt_base_context_obs_learned_dominance_context_success_envclip_train_to_test/eval.log +3 -0
  41. workspace/runs/ctt_base_context_obs_learned_dominance_context_success_envclip_train_to_test/git_hash.txt +1 -0
  42. workspace/runs/ctt_base_context_obs_learned_dominance_context_success_envclip_train_to_test/metrics.json +0 -0
  43. workspace/runs/ctt_base_context_obs_learned_dominance_context_success_envclip_train_to_test/metrics_by_seed.json +2762 -0
  44. workspace/runs/ctt_base_context_obs_learned_dominance_context_success_envclip_train_to_test/metrics_by_task.json +117 -0
  45. workspace/runs/ctt_base_context_obs_learned_dominance_context_success_envclip_train_to_test/report.md +17 -0
  46. workspace/runs/ctt_base_context_obs_learned_dominance_context_success_envclip_train_to_test/split_hash.txt +14 -0
  47. workspace/runs/ctt_base_context_obs_learned_dominance_context_success_envclip_train_to_test/table.tex +8 -0
  48. workspace/runs/ctt_base_context_obs_learned_dominance_context_success_envclip_train_to_test/train.log +4 -0
  49. workspace/runs/ctt_base_context_obs_learned_dominance_context_tangent_envclip_train_to_test/command.txt +1 -0
  50. workspace/runs/ctt_base_context_obs_learned_dominance_context_tangent_envclip_train_to_test/config.yaml +11 -0
workspace/runs/ctt_base_context_obs_dominance_envclip_train_to_test/command.txt ADDED
@@ -0,0 +1 @@
 
 
1
+ python scripts/eval_dominance_selector.py --calibration-input runs/ctt_base_context_obs_train_cal_envclip_rollout_comparison/combined_measured_candidates.json --calibration-target-index data/cil_charts_rgb_refs/train/index.json --eval-input runs/ctt_base_context_obs_test_envclip_rollout_comparison/combined_measured_candidates.json --eval-target-index data/cil_charts_rgb_refs/test/index.json --checkpoint-template runs/ctt_residual_base_context_obs_seed{seed}/model.pt --score-source row --out-dir runs/ctt_base_context_obs_dominance_envclip_train_to_test --k 8 --bootstrap-samples 1000
workspace/runs/ctt_base_context_obs_dominance_envclip_train_to_test/config.yaml ADDED
@@ -0,0 +1,11 @@
 
 
 
 
 
 
 
 
 
 
 
 
1
+ alpha: 0.1
2
+ bootstrap_samples: 1000
3
+ calibration_input: runs/ctt_base_context_obs_train_cal_envclip_rollout_comparison/combined_measured_candidates.json
4
+ calibration_target_index: data/cil_charts_rgb_refs/train/index.json
5
+ checkpoint_template: runs/ctt_residual_base_context_obs_seed{seed}/model.pt
6
+ eval_input: runs/ctt_base_context_obs_test_envclip_rollout_comparison/combined_measured_candidates.json
7
+ eval_target_index: data/cil_charts_rgb_refs/test/index.json
8
+ k: 8
9
+ out_dir: runs/ctt_base_context_obs_dominance_envclip_train_to_test
10
+ score_source: row
11
+ tau: auto
workspace/runs/ctt_base_context_obs_dominance_envclip_train_to_test/data_hash.txt ADDED
@@ -0,0 +1,6 @@
 
 
 
 
 
 
 
1
+ {
2
+ "calibration_input": "06375099001e740a377c402e34b77bf6974be3dc2a468974340ce41eb1b487f0",
3
+ "calibration_target_index": "67973c50072aa87a3b97533b6292b08a8b4d88a95571dc22e91b6e8fd19e0d85",
4
+ "eval_input": "57df43ff4802379dfd60b5209d5a04028d721256406685decededccdbec0d49c",
5
+ "eval_target_index": "dee1b5225485df6624747dd4c5074bcf514685b54017d05e698d2051638fbf36"
6
+ }
workspace/runs/ctt_base_context_obs_dominance_envclip_train_to_test/eval.log ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ evaluated calibrated fallback rule on held-out measured rollout rows
2
+ eval_input=runs/ctt_base_context_obs_test_envclip_rollout_comparison/combined_measured_candidates.json
3
+ num_eval_rows=144
workspace/runs/ctt_base_context_obs_dominance_envclip_train_to_test/git_hash.txt ADDED
@@ -0,0 +1 @@
 
 
1
+ 14e81292e2004925f81e0bb4810f6f144bfd87b7
workspace/runs/ctt_base_context_obs_dominance_envclip_train_to_test/metrics.json ADDED
The diff for this file is too large to render. See raw diff
 
workspace/runs/ctt_base_context_obs_dominance_envclip_train_to_test/metrics_by_seed.json ADDED
The diff for this file is too large to render. See raw diff
 
workspace/runs/ctt_base_context_obs_dominance_envclip_train_to_test/metrics_by_task.json ADDED
@@ -0,0 +1,157 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "LiftPegUpright-v1": {
3
+ "base_predicted_score": 3.8269843657811484,
4
+ "base_success": 0.2,
5
+ "base_utility": 0.9498705387115478,
6
+ "branch_car": 0.4276172121365865,
7
+ "coverage": 0.0,
8
+ "execute_generated": 0.0,
9
+ "fallback_rate": 1.0,
10
+ "fallback_to_base": 1.0,
11
+ "hidden_chart_oracle_success": 1.0,
12
+ "hidden_chart_oracle_utility": 2.0,
13
+ "lcb_margin": -4.828884870310625,
14
+ "measured_margin": -0.29843822717666624,
15
+ "outcome_ptr": 0.4666666666666667,
16
+ "predicted_margin": 1.045054308573405,
17
+ "proposal_oracle_success": 0.4666666666666667,
18
+ "proposal_oracle_utility": 1.2244940986235937,
19
+ "selected_success": 0.2,
20
+ "selected_success_gain_over_base": 0.0,
21
+ "selected_utility": 0.9498705387115478,
22
+ "selected_utility_gain_over_base": 0.0,
23
+ "selector_regret": 0.4276172121365865,
24
+ "success_selector_gap": 0.3333333333333333,
25
+ "success_support_gap": 0.5333333333333333,
26
+ "success_total_car_to_hidden": 0.8,
27
+ "support_gap": 0.7755059013764064,
28
+ "top_generated_success": 0.13333333333333333,
29
+ "top_generated_utility": 0.6514323115348816,
30
+ "top_index": 0.0,
31
+ "top_predicted_score": 4.872038674354553
32
+ },
33
+ "PickCube-v1": {
34
+ "base_predicted_score": 5.362143039703369,
35
+ "base_success": 0.0,
36
+ "base_utility": 0.5839591026306152,
37
+ "branch_car": 0.0,
38
+ "coverage": 0.0,
39
+ "execute_generated": 0.0,
40
+ "fallback_rate": 1.0,
41
+ "fallback_to_base": 1.0,
42
+ "hidden_chart_oracle_success": 1.0,
43
+ "hidden_chart_oracle_utility": 2.0,
44
+ "lcb_margin": -5.7723739768068,
45
+ "measured_margin": -0.47562045852343243,
46
+ "outcome_ptr": 0.0,
47
+ "predicted_margin": 0.10156520207722981,
48
+ "proposal_oracle_success": 0.0,
49
+ "proposal_oracle_utility": 0.13425676027933756,
50
+ "selected_success": 0.0,
51
+ "selected_success_gain_over_base": 0.0,
52
+ "selected_utility": 0.5839591026306152,
53
+ "selected_utility_gain_over_base": 0.0,
54
+ "selector_regret": 0.0,
55
+ "success_selector_gap": 0.0,
56
+ "success_support_gap": 1.0,
57
+ "success_total_car_to_hidden": 1.0,
58
+ "support_gap": 1.8657432397206624,
59
+ "top_generated_success": 0.0,
60
+ "top_generated_utility": 0.10833864410718282,
61
+ "top_index": 0.0,
62
+ "top_predicted_score": 5.463708241780599
63
+ },
64
+ "PullCube-v1": {
65
+ "base_predicted_score": 2.447210181752841,
66
+ "base_success": 0.36,
67
+ "base_utility": 0.7751704612234608,
68
+ "branch_car": 0.391817180281505,
69
+ "coverage": 0.14666666666666667,
70
+ "execute_generated": 0.14666666666666667,
71
+ "fallback_rate": 0.8533333333333334,
72
+ "fallback_to_base": 0.8533333333333334,
73
+ "hidden_chart_oracle_success": 0.52,
74
+ "hidden_chart_oracle_utility": 1.1467878860235214,
75
+ "lcb_margin": -3.2647068614761037,
76
+ "measured_margin": -0.1983167066782092,
77
+ "outcome_ptr": 0.44,
78
+ "predicted_margin": 2.6092323174079257,
79
+ "proposal_oracle_success": 0.4666666666666667,
80
+ "proposal_oracle_utility": 0.9816135169876119,
81
+ "selected_success": 0.3466666666666667,
82
+ "selected_success_gain_over_base": -0.013333333333333334,
83
+ "selected_utility": 0.7469129716325551,
84
+ "selected_utility_gain_over_base": -0.028257489590905606,
85
+ "selector_regret": 0.391817180281505,
86
+ "success_selector_gap": 0.2,
87
+ "success_support_gap": 0.09333333333333334,
88
+ "success_total_car_to_hidden": 0.24,
89
+ "support_gap": 0.23527305972451965,
90
+ "top_generated_success": 0.26666666666666666,
91
+ "top_generated_utility": 0.5768537545452516,
92
+ "top_index": 0.0,
93
+ "top_predicted_score": 5.056442499160767
94
+ },
95
+ "PushCube-v1": {
96
+ "base_predicted_score": 2.019581354326672,
97
+ "base_success": 0.16666666666666666,
98
+ "base_utility": 0.3924041787783305,
99
+ "branch_car": 0.8607266768813133,
100
+ "coverage": 0.2222222222222222,
101
+ "execute_generated": 0.2222222222222222,
102
+ "fallback_rate": 0.7777777777777778,
103
+ "fallback_to_base": 0.7777777777777778,
104
+ "hidden_chart_oracle_success": 0.8333333333333334,
105
+ "hidden_chart_oracle_utility": 1.7015800600250561,
106
+ "lcb_margin": -1.9365553119116359,
107
+ "measured_margin": 0.22239751203192604,
108
+ "outcome_ptr": 0.7777777777777778,
109
+ "predicted_margin": 3.9373838669723935,
110
+ "proposal_oracle_success": 0.6666666666666666,
111
+ "proposal_oracle_utility": 1.365773618221283,
112
+ "selected_success": 0.2222222222222222,
113
+ "selected_success_gain_over_base": 0.05555555555555555,
114
+ "selected_utility": 0.5065385094947286,
115
+ "selected_utility_gain_over_base": 0.11413433071639803,
116
+ "selector_regret": 0.8607266768813133,
117
+ "success_selector_gap": 0.4444444444444444,
118
+ "success_support_gap": 0.3333333333333333,
119
+ "success_total_car_to_hidden": 0.6111111111111112,
120
+ "support_gap": 0.634226381778717,
121
+ "top_generated_success": 0.2777777777777778,
122
+ "top_generated_utility": 0.6148016908102565,
123
+ "top_index": 0.0,
124
+ "top_predicted_score": 5.956965221299066
125
+ },
126
+ "StackCube-v1": {
127
+ "base_predicted_score": 5.208852457277702,
128
+ "base_success": 0.2727272727272727,
129
+ "base_utility": 0.7763947045261209,
130
+ "branch_car": 0.2826216333743298,
131
+ "coverage": 0.0,
132
+ "execute_generated": 0.0,
133
+ "fallback_rate": 1.0,
134
+ "fallback_to_base": 1.0,
135
+ "hidden_chart_oracle_success": 1.0,
136
+ "hidden_chart_oracle_utility": 2.0,
137
+ "lcb_margin": -4.981105696748603,
138
+ "measured_margin": -0.2091979912736199,
139
+ "outcome_ptr": 0.5757575757575758,
140
+ "predicted_margin": 0.8928334821354259,
141
+ "proposal_oracle_success": 0.45454545454545453,
142
+ "proposal_oracle_utility": 1.0546941955884297,
143
+ "selected_success": 0.2727272727272727,
144
+ "selected_success_gain_over_base": 0.0,
145
+ "selected_utility": 0.7763947045261209,
146
+ "selected_utility_gain_over_base": 0.0,
147
+ "selector_regret": 0.2826216333743298,
148
+ "success_selector_gap": 0.18181818181818182,
149
+ "success_support_gap": 0.5454545454545454,
150
+ "success_total_car_to_hidden": 0.7272727272727273,
151
+ "support_gap": 0.9453058044115702,
152
+ "top_generated_success": 0.21212121212121213,
153
+ "top_generated_utility": 0.5671967132525011,
154
+ "top_index": 0.0,
155
+ "top_predicted_score": 6.101685939413128
156
+ }
157
+ }
workspace/runs/ctt_base_context_obs_dominance_envclip_train_to_test/report.md ADDED
@@ -0,0 +1,16 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ # Dominance-Calibrated CTT Selector
2
+
3
+ Calibration rows: `432`
4
+ Eval rows: `144`
5
+ Alpha: `0.1`
6
+ Residual quantile: `5.873939`
7
+ Tau: `-0.168926` (`auto`)
8
+
9
+ The threshold is fit on calibration rows only. Eval outcomes are used only for reporting.
10
+
11
+ | Split | Coverage | Fallback | Base success | Selected success | Proposal oracle | OutcomePTR | Success support gap | Success selector gap |
12
+ | --- | ---: | ---: | ---: | ---: | ---: | ---: | ---: | ---: |
13
+ | calibration | 0.0949 | 0.9051 | 0.2778 | 0.2986 | 0.4722 | 0.4907 | 0.2315 | 0.2361 |
14
+ | eval | 0.1042 | 0.8958 | 0.2917 | 0.2917 | 0.4792 | 0.5069 | 0.2917 | 0.2361 |
15
+
16
+ This is a calibrated fallback diagnostic. It is not a final safety claim because unsafe-contact labels are not measured yet.
workspace/runs/ctt_base_context_obs_dominance_envclip_train_to_test/split_hash.txt ADDED
@@ -0,0 +1,14 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "calibration_target_index": {
3
+ "content_hash": "67f7b4a692f7c7e71da24378dc71e3399c5d35afc14c494869eba6087b765c42",
4
+ "retrieval_index_allowed": true,
5
+ "split": "train",
6
+ "split_hash": "867a94a1c67686c082765671a7bf8d74a9d633085e1b2619c83197c60bf58372"
7
+ },
8
+ "eval_target_index": {
9
+ "content_hash": "0c38d73dc9bc4ab756e12006671b5d12f1a23358b281e54b3f6a9ae3cbf6e9af",
10
+ "retrieval_index_allowed": false,
11
+ "split": "test",
12
+ "split_hash": "84151df9e88c219b76751677558cc5bbbc19082bf358cea7330158e9e403e25f"
13
+ }
14
+ }
workspace/runs/ctt_base_context_obs_dominance_envclip_train_to_test/table.tex ADDED
@@ -0,0 +1,8 @@
 
 
 
 
 
 
 
 
 
1
+ % Auto-generated by scripts/eval_dominance_selector.py
2
+ \begin{tabular}{lrrrrrrrr}
3
+ \toprule
4
+ Rows & Coverage & Fallback & Base succ. & Selected succ. & Oracle succ. & OutcomePTR & Succ. support gap & Succ. selector gap \\
5
+ \midrule
6
+ 144 & 0.1042 & 0.8958 & 0.2917 & 0.2917 & 0.4792 & 0.5069 & 0.2917 & 0.2361 \\
7
+ \bottomrule
8
+ \end{tabular}
workspace/runs/ctt_base_context_obs_dominance_envclip_train_to_test/train.log ADDED
@@ -0,0 +1,4 @@
 
 
 
 
 
1
+ fit conformal residual quantile and tau on calibration measured rows only
2
+ calibration_input=runs/ctt_base_context_obs_train_cal_envclip_rollout_comparison/combined_measured_candidates.json
3
+ residual_quantile=5.873939
4
+ tau=-0.168926
workspace/runs/ctt_base_context_obs_learned_dominance_basic_envclip_train_to_test/command.txt ADDED
@@ -0,0 +1 @@
 
 
1
+ python scripts/eval_learned_dominance_selector.py --calibration-input runs/ctt_base_context_obs_train_cal_envclip_rollout_comparison/combined_measured_candidates.json --calibration-target-index data/cil_charts_rgb_refs/train/index.json --eval-input runs/ctt_base_context_obs_test_envclip_rollout_comparison/combined_measured_candidates.json --eval-target-index data/cil_charts_rgb_refs/test/index.json --checkpoint-template runs/ctt_residual_base_context_obs_seed{seed}/model.pt --out-dir runs/ctt_base_context_obs_learned_dominance_basic_envclip_train_to_test --feature-set basic --target utility_margin --k 8 --bootstrap-samples 1000
workspace/runs/ctt_base_context_obs_learned_dominance_basic_envclip_train_to_test/config.yaml ADDED
@@ -0,0 +1,11 @@
 
 
 
 
 
 
 
 
 
 
 
 
1
+ bootstrap_samples: 1000
2
+ calibration_input: runs/ctt_base_context_obs_train_cal_envclip_rollout_comparison/combined_measured_candidates.json
3
+ calibration_target_index: data/cil_charts_rgb_refs/train/index.json
4
+ checkpoint_template: runs/ctt_residual_base_context_obs_seed{seed}/model.pt
5
+ eval_input: runs/ctt_base_context_obs_test_envclip_rollout_comparison/combined_measured_candidates.json
6
+ eval_target_index: data/cil_charts_rgb_refs/test/index.json
7
+ feature_set: basic
8
+ k: 8
9
+ out_dir: runs/ctt_base_context_obs_learned_dominance_basic_envclip_train_to_test
10
+ ridge_lambdas: 0,0.01,0.1,1,10,100
11
+ target: utility_margin
workspace/runs/ctt_base_context_obs_learned_dominance_basic_envclip_train_to_test/data_hash.txt ADDED
@@ -0,0 +1,6 @@
 
 
 
 
 
 
 
1
+ {
2
+ "calibration_input": "06375099001e740a377c402e34b77bf6974be3dc2a468974340ce41eb1b487f0",
3
+ "calibration_target_index": "67973c50072aa87a3b97533b6292b08a8b4d88a95571dc22e91b6e8fd19e0d85",
4
+ "eval_input": "57df43ff4802379dfd60b5209d5a04028d721256406685decededccdbec0d49c",
5
+ "eval_target_index": "dee1b5225485df6624747dd4c5074bcf514685b54017d05e698d2051638fbf36"
6
+ }
workspace/runs/ctt_base_context_obs_learned_dominance_basic_envclip_train_to_test/eval.log ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ evaluated learned fallback rule on held-out measured rollout rows
2
+ eval_input=runs/ctt_base_context_obs_test_envclip_rollout_comparison/combined_measured_candidates.json
3
+ num_eval_rows=144
workspace/runs/ctt_base_context_obs_learned_dominance_basic_envclip_train_to_test/git_hash.txt ADDED
@@ -0,0 +1 @@
 
 
1
+ 14e81292e2004925f81e0bb4810f6f144bfd87b7
workspace/runs/ctt_base_context_obs_learned_dominance_basic_envclip_train_to_test/metrics.json ADDED
The diff for this file is too large to render. See raw diff
 
workspace/runs/ctt_base_context_obs_learned_dominance_basic_envclip_train_to_test/metrics_by_seed.json ADDED
@@ -0,0 +1,2762 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "1073161778/train0": {
3
+ "base_success": 1.0,
4
+ "base_utility": 2.0,
5
+ "coverage": 0.0,
6
+ "execute_generated": 0.0,
7
+ "fallback_rate": 1.0,
8
+ "hidden_chart_oracle_success": 1.0,
9
+ "hidden_chart_oracle_utility": 2.0,
10
+ "outcome_ptr": 0.0,
11
+ "predicted_margin": 0.08246751297046068,
12
+ "proposal_oracle_success": 1.0,
13
+ "proposal_oracle_utility": 2.0,
14
+ "selected_candidate_index": 4.0,
15
+ "selected_success": 1.0,
16
+ "selected_success_gain_over_base": 0.0,
17
+ "selected_utility": 2.0,
18
+ "selected_utility_gain_over_base": 0.0,
19
+ "selector_regret": 0.0,
20
+ "success_selector_gap": 0.0,
21
+ "success_support_gap": 0.0,
22
+ "support_gap": 0.0,
23
+ "tau": 0.10322580228482307
24
+ },
25
+ "1073161778/train1": {
26
+ "base_success": 1.0,
27
+ "base_utility": 2.0,
28
+ "coverage": 0.0,
29
+ "execute_generated": 0.0,
30
+ "fallback_rate": 1.0,
31
+ "hidden_chart_oracle_success": 1.0,
32
+ "hidden_chart_oracle_utility": 2.0,
33
+ "outcome_ptr": 0.0,
34
+ "predicted_margin": -0.004279528949672443,
35
+ "proposal_oracle_success": 0.0,
36
+ "proposal_oracle_utility": 0.7071996331214905,
37
+ "selected_candidate_index": 3.0,
38
+ "selected_success": 1.0,
39
+ "selected_success_gain_over_base": 0.0,
40
+ "selected_utility": 2.0,
41
+ "selected_utility_gain_over_base": 0.0,
42
+ "selector_regret": 0.0,
43
+ "success_selector_gap": 0.0,
44
+ "success_support_gap": 1.0,
45
+ "support_gap": 1.2928003668785095,
46
+ "tau": 0.10322580228482307
47
+ },
48
+ "1073161778/train2": {
49
+ "base_success": 1.0,
50
+ "base_utility": 2.0,
51
+ "coverage": 0.0,
52
+ "execute_generated": 0.0,
53
+ "fallback_rate": 1.0,
54
+ "hidden_chart_oracle_success": 1.0,
55
+ "hidden_chart_oracle_utility": 2.0,
56
+ "outcome_ptr": 0.0,
57
+ "predicted_margin": 0.08985900631683494,
58
+ "proposal_oracle_success": 1.0,
59
+ "proposal_oracle_utility": 2.0,
60
+ "selected_candidate_index": 2.0,
61
+ "selected_success": 1.0,
62
+ "selected_success_gain_over_base": 0.0,
63
+ "selected_utility": 2.0,
64
+ "selected_utility_gain_over_base": 0.0,
65
+ "selector_regret": 0.0,
66
+ "success_selector_gap": 0.0,
67
+ "success_support_gap": 0.0,
68
+ "support_gap": 0.0,
69
+ "tau": 0.10322580228482307
70
+ },
71
+ "1113371261/train0": {
72
+ "base_success": 0.0,
73
+ "base_utility": 0.15842342376708984,
74
+ "coverage": 0.0,
75
+ "execute_generated": 0.0,
76
+ "fallback_rate": 1.0,
77
+ "hidden_chart_oracle_success": 1.0,
78
+ "hidden_chart_oracle_utility": 2.0,
79
+ "outcome_ptr": 0.0,
80
+ "predicted_margin": -0.03658029791986532,
81
+ "proposal_oracle_success": 0.0,
82
+ "proposal_oracle_utility": 0.10274861752986908,
83
+ "selected_candidate_index": 1.0,
84
+ "selected_success": 0.0,
85
+ "selected_success_gain_over_base": 0.0,
86
+ "selected_utility": 0.15842342376708984,
87
+ "selected_utility_gain_over_base": 0.0,
88
+ "selector_regret": 0.0,
89
+ "success_selector_gap": 0.0,
90
+ "success_support_gap": 1.0,
91
+ "support_gap": 1.897251382470131,
92
+ "tau": 0.10322580228482307
93
+ },
94
+ "1113371261/train1": {
95
+ "base_success": 0.0,
96
+ "base_utility": 0.15842342376708984,
97
+ "coverage": 0.0,
98
+ "execute_generated": 0.0,
99
+ "fallback_rate": 1.0,
100
+ "hidden_chart_oracle_success": 1.0,
101
+ "hidden_chart_oracle_utility": 2.0,
102
+ "outcome_ptr": 1.0,
103
+ "predicted_margin": 0.024652380855946857,
104
+ "proposal_oracle_success": 0.0,
105
+ "proposal_oracle_utility": 0.18093222379684448,
106
+ "selected_candidate_index": 2.0,
107
+ "selected_success": 0.0,
108
+ "selected_success_gain_over_base": 0.0,
109
+ "selected_utility": 0.15842342376708984,
110
+ "selected_utility_gain_over_base": 0.0,
111
+ "selector_regret": 0.02250880002975464,
112
+ "success_selector_gap": 0.0,
113
+ "success_support_gap": 1.0,
114
+ "support_gap": 1.8190677762031555,
115
+ "tau": 0.10322580228482307
116
+ },
117
+ "1113371261/train2": {
118
+ "base_success": 0.0,
119
+ "base_utility": 0.15842342376708984,
120
+ "coverage": 0.0,
121
+ "execute_generated": 0.0,
122
+ "fallback_rate": 1.0,
123
+ "hidden_chart_oracle_success": 1.0,
124
+ "hidden_chart_oracle_utility": 2.0,
125
+ "outcome_ptr": 1.0,
126
+ "predicted_margin": -0.04922313844967605,
127
+ "proposal_oracle_success": 0.0,
128
+ "proposal_oracle_utility": 0.19527167081832886,
129
+ "selected_candidate_index": 4.0,
130
+ "selected_success": 0.0,
131
+ "selected_success_gain_over_base": 0.0,
132
+ "selected_utility": 0.15842342376708984,
133
+ "selected_utility_gain_over_base": 0.0,
134
+ "selector_regret": 0.036848247051239014,
135
+ "success_selector_gap": 0.0,
136
+ "success_support_gap": 1.0,
137
+ "support_gap": 1.8047283291816711,
138
+ "tau": 0.10322580228482307
139
+ },
140
+ "1151410346/train0": {
141
+ "base_success": 0.0,
142
+ "base_utility": 0.2260012924671173,
143
+ "coverage": 0.0,
144
+ "execute_generated": 0.0,
145
+ "fallback_rate": 1.0,
146
+ "hidden_chart_oracle_success": 1.0,
147
+ "hidden_chart_oracle_utility": 2.0,
148
+ "outcome_ptr": 1.0,
149
+ "predicted_margin": 0.09485503943952373,
150
+ "proposal_oracle_success": 0.0,
151
+ "proposal_oracle_utility": 0.8663521409034729,
152
+ "selected_candidate_index": 5.0,
153
+ "selected_success": 0.0,
154
+ "selected_success_gain_over_base": 0.0,
155
+ "selected_utility": 0.2260012924671173,
156
+ "selected_utility_gain_over_base": 0.0,
157
+ "selector_regret": 0.6403508484363556,
158
+ "success_selector_gap": 0.0,
159
+ "success_support_gap": 1.0,
160
+ "support_gap": 1.133647859096527,
161
+ "tau": 0.10322580228482307
162
+ },
163
+ "1151410346/train1": {
164
+ "base_success": 0.0,
165
+ "base_utility": 0.2260012924671173,
166
+ "coverage": 1.0,
167
+ "execute_generated": 1.0,
168
+ "fallback_rate": 0.0,
169
+ "hidden_chart_oracle_success": 1.0,
170
+ "hidden_chart_oracle_utility": 2.0,
171
+ "outcome_ptr": 1.0,
172
+ "predicted_margin": 0.10729885909199958,
173
+ "proposal_oracle_success": 0.0,
174
+ "proposal_oracle_utility": 0.8138919472694397,
175
+ "selected_candidate_index": 1.0,
176
+ "selected_success": 0.0,
177
+ "selected_success_gain_over_base": 0.0,
178
+ "selected_utility": 0.812879204750061,
179
+ "selected_utility_gain_over_base": 0.5868779122829437,
180
+ "selector_regret": 0.0010127425193786621,
181
+ "success_selector_gap": 0.0,
182
+ "success_support_gap": 1.0,
183
+ "support_gap": 1.1861080527305603,
184
+ "tau": 0.10322580228482307
185
+ },
186
+ "1151410346/train2": {
187
+ "base_success": 0.0,
188
+ "base_utility": 0.2260012924671173,
189
+ "coverage": 1.0,
190
+ "execute_generated": 1.0,
191
+ "fallback_rate": 0.0,
192
+ "hidden_chart_oracle_success": 1.0,
193
+ "hidden_chart_oracle_utility": 2.0,
194
+ "outcome_ptr": 0.0,
195
+ "predicted_margin": 0.11097800106606831,
196
+ "proposal_oracle_success": 0.0,
197
+ "proposal_oracle_utility": 0.21912741661071777,
198
+ "selected_candidate_index": 4.0,
199
+ "selected_success": 0.0,
200
+ "selected_success_gain_over_base": 0.0,
201
+ "selected_utility": 0.2044079303741455,
202
+ "selected_utility_gain_over_base": -0.021593362092971802,
203
+ "selector_regret": 0.014719486236572266,
204
+ "success_selector_gap": 0.0,
205
+ "success_support_gap": 1.0,
206
+ "support_gap": 1.7808725833892822,
207
+ "tau": 0.10322580228482307
208
+ },
209
+ "1261930456/train0": {
210
+ "base_success": 1.0,
211
+ "base_utility": 2.0,
212
+ "coverage": 1.0,
213
+ "execute_generated": 1.0,
214
+ "fallback_rate": 0.0,
215
+ "hidden_chart_oracle_success": 1.0,
216
+ "hidden_chart_oracle_utility": 2.0,
217
+ "outcome_ptr": 0.0,
218
+ "predicted_margin": 0.14109172645301515,
219
+ "proposal_oracle_success": 1.0,
220
+ "proposal_oracle_utility": 2.0,
221
+ "selected_candidate_index": 3.0,
222
+ "selected_success": 0.0,
223
+ "selected_success_gain_over_base": -1.0,
224
+ "selected_utility": 0.06716175377368927,
225
+ "selected_utility_gain_over_base": -1.9328382462263107,
226
+ "selector_regret": 1.9328382462263107,
227
+ "success_selector_gap": 1.0,
228
+ "success_support_gap": 0.0,
229
+ "support_gap": 0.0,
230
+ "tau": 0.10322580228482307
231
+ },
232
+ "1261930456/train1": {
233
+ "base_success": 1.0,
234
+ "base_utility": 2.0,
235
+ "coverage": 1.0,
236
+ "execute_generated": 1.0,
237
+ "fallback_rate": 0.0,
238
+ "hidden_chart_oracle_success": 1.0,
239
+ "hidden_chart_oracle_utility": 2.0,
240
+ "outcome_ptr": 0.0,
241
+ "predicted_margin": 0.18477824382557312,
242
+ "proposal_oracle_success": 1.0,
243
+ "proposal_oracle_utility": 2.0,
244
+ "selected_candidate_index": 2.0,
245
+ "selected_success": 1.0,
246
+ "selected_success_gain_over_base": 0.0,
247
+ "selected_utility": 2.0,
248
+ "selected_utility_gain_over_base": 0.0,
249
+ "selector_regret": 0.0,
250
+ "success_selector_gap": 0.0,
251
+ "success_support_gap": 0.0,
252
+ "support_gap": 0.0,
253
+ "tau": 0.10322580228482307
254
+ },
255
+ "1261930456/train2": {
256
+ "base_success": 1.0,
257
+ "base_utility": 2.0,
258
+ "coverage": 1.0,
259
+ "execute_generated": 1.0,
260
+ "fallback_rate": 0.0,
261
+ "hidden_chart_oracle_success": 1.0,
262
+ "hidden_chart_oracle_utility": 2.0,
263
+ "outcome_ptr": 0.0,
264
+ "predicted_margin": 0.15416333590590994,
265
+ "proposal_oracle_success": 1.0,
266
+ "proposal_oracle_utility": 2.0,
267
+ "selected_candidate_index": 4.0,
268
+ "selected_success": 0.0,
269
+ "selected_success_gain_over_base": -1.0,
270
+ "selected_utility": 0.08501242101192474,
271
+ "selected_utility_gain_over_base": -1.9149875789880753,
272
+ "selector_regret": 1.9149875789880753,
273
+ "success_selector_gap": 1.0,
274
+ "success_support_gap": 0.0,
275
+ "support_gap": 0.0,
276
+ "tau": 0.10322580228482307
277
+ },
278
+ "1324133122/train0": {
279
+ "base_success": 0.0,
280
+ "base_utility": 0.09220491349697113,
281
+ "coverage": 0.0,
282
+ "execute_generated": 0.0,
283
+ "fallback_rate": 1.0,
284
+ "hidden_chart_oracle_success": 1.0,
285
+ "hidden_chart_oracle_utility": 2.0,
286
+ "outcome_ptr": 0.0,
287
+ "predicted_margin": -0.10253807599404008,
288
+ "proposal_oracle_success": 0.0,
289
+ "proposal_oracle_utility": 0.054989203810691833,
290
+ "selected_candidate_index": 2.0,
291
+ "selected_success": 0.0,
292
+ "selected_success_gain_over_base": 0.0,
293
+ "selected_utility": 0.09220491349697113,
294
+ "selected_utility_gain_over_base": 0.0,
295
+ "selector_regret": 0.0,
296
+ "success_selector_gap": 0.0,
297
+ "success_support_gap": 1.0,
298
+ "support_gap": 1.9450107961893082,
299
+ "tau": 0.10322580228482307
300
+ },
301
+ "1324133122/train1": {
302
+ "base_success": 0.0,
303
+ "base_utility": 0.09220491349697113,
304
+ "coverage": 0.0,
305
+ "execute_generated": 0.0,
306
+ "fallback_rate": 1.0,
307
+ "hidden_chart_oracle_success": 1.0,
308
+ "hidden_chart_oracle_utility": 2.0,
309
+ "outcome_ptr": 1.0,
310
+ "predicted_margin": 0.037668973850787976,
311
+ "proposal_oracle_success": 0.0,
312
+ "proposal_oracle_utility": 0.1324586570262909,
313
+ "selected_candidate_index": 6.0,
314
+ "selected_success": 0.0,
315
+ "selected_success_gain_over_base": 0.0,
316
+ "selected_utility": 0.09220491349697113,
317
+ "selected_utility_gain_over_base": 0.0,
318
+ "selector_regret": 0.04025374352931976,
319
+ "success_selector_gap": 0.0,
320
+ "success_support_gap": 1.0,
321
+ "support_gap": 1.867541342973709,
322
+ "tau": 0.10322580228482307
323
+ },
324
+ "1324133122/train2": {
325
+ "base_success": 0.0,
326
+ "base_utility": 0.09220491349697113,
327
+ "coverage": 0.0,
328
+ "execute_generated": 0.0,
329
+ "fallback_rate": 1.0,
330
+ "hidden_chart_oracle_success": 1.0,
331
+ "hidden_chart_oracle_utility": 2.0,
332
+ "outcome_ptr": 1.0,
333
+ "predicted_margin": -0.051176984261192694,
334
+ "proposal_oracle_success": 0.0,
335
+ "proposal_oracle_utility": 0.19291165471076965,
336
+ "selected_candidate_index": 2.0,
337
+ "selected_success": 0.0,
338
+ "selected_success_gain_over_base": 0.0,
339
+ "selected_utility": 0.09220491349697113,
340
+ "selected_utility_gain_over_base": 0.0,
341
+ "selector_regret": 0.10070674121379852,
342
+ "success_selector_gap": 0.0,
343
+ "success_support_gap": 1.0,
344
+ "support_gap": 1.8070883452892303,
345
+ "tau": 0.10322580228482307
346
+ },
347
+ "1324642411/train0": {
348
+ "base_success": 0.0,
349
+ "base_utility": 0.10647052526473999,
350
+ "coverage": 1.0,
351
+ "execute_generated": 1.0,
352
+ "fallback_rate": 0.0,
353
+ "hidden_chart_oracle_success": 1.0,
354
+ "hidden_chart_oracle_utility": 2.0,
355
+ "outcome_ptr": 1.0,
356
+ "predicted_margin": 0.2286022589394483,
357
+ "proposal_oracle_success": 0.0,
358
+ "proposal_oracle_utility": 0.17193320393562317,
359
+ "selected_candidate_index": 3.0,
360
+ "selected_success": 0.0,
361
+ "selected_success_gain_over_base": 0.0,
362
+ "selected_utility": 0.08104583621025085,
363
+ "selected_utility_gain_over_base": -0.025424689054489136,
364
+ "selector_regret": 0.09088736772537231,
365
+ "success_selector_gap": 0.0,
366
+ "success_support_gap": 1.0,
367
+ "support_gap": 1.8280667960643768,
368
+ "tau": 0.10322580228482307
369
+ },
370
+ "1324642411/train1": {
371
+ "base_success": 0.0,
372
+ "base_utility": 0.10647052526473999,
373
+ "coverage": 1.0,
374
+ "execute_generated": 1.0,
375
+ "fallback_rate": 0.0,
376
+ "hidden_chart_oracle_success": 1.0,
377
+ "hidden_chart_oracle_utility": 2.0,
378
+ "outcome_ptr": 1.0,
379
+ "predicted_margin": 0.14601854127213684,
380
+ "proposal_oracle_success": 0.0,
381
+ "proposal_oracle_utility": 0.1836909055709839,
382
+ "selected_candidate_index": 1.0,
383
+ "selected_success": 0.0,
384
+ "selected_success_gain_over_base": 0.0,
385
+ "selected_utility": 0.10393516719341278,
386
+ "selected_utility_gain_over_base": -0.0025353580713272095,
387
+ "selector_regret": 0.0797557383775711,
388
+ "success_selector_gap": 0.0,
389
+ "success_support_gap": 1.0,
390
+ "support_gap": 1.8163090944290161,
391
+ "tau": 0.10322580228482307
392
+ },
393
+ "1324642411/train2": {
394
+ "base_success": 0.0,
395
+ "base_utility": 0.10647052526473999,
396
+ "coverage": 0.0,
397
+ "execute_generated": 0.0,
398
+ "fallback_rate": 1.0,
399
+ "hidden_chart_oracle_success": 1.0,
400
+ "hidden_chart_oracle_utility": 2.0,
401
+ "outcome_ptr": 1.0,
402
+ "predicted_margin": -0.14243102495968665,
403
+ "proposal_oracle_success": 1.0,
404
+ "proposal_oracle_utility": 2.0,
405
+ "selected_candidate_index": 2.0,
406
+ "selected_success": 0.0,
407
+ "selected_success_gain_over_base": 0.0,
408
+ "selected_utility": 0.10647052526473999,
409
+ "selected_utility_gain_over_base": 0.0,
410
+ "selector_regret": 1.89352947473526,
411
+ "success_selector_gap": 1.0,
412
+ "success_support_gap": 0.0,
413
+ "support_gap": 0.0,
414
+ "tau": 0.10322580228482307
415
+ },
416
+ "1375492786/train0": {
417
+ "base_success": 1.0,
418
+ "base_utility": 2.0,
419
+ "coverage": 0.0,
420
+ "execute_generated": 0.0,
421
+ "fallback_rate": 1.0,
422
+ "hidden_chart_oracle_success": 1.0,
423
+ "hidden_chart_oracle_utility": 2.0,
424
+ "outcome_ptr": 0.0,
425
+ "predicted_margin": -0.17629684258001238,
426
+ "proposal_oracle_success": 1.0,
427
+ "proposal_oracle_utility": 2.0,
428
+ "selected_candidate_index": 7.0,
429
+ "selected_success": 1.0,
430
+ "selected_success_gain_over_base": 0.0,
431
+ "selected_utility": 2.0,
432
+ "selected_utility_gain_over_base": 0.0,
433
+ "selector_regret": 0.0,
434
+ "success_selector_gap": 0.0,
435
+ "success_support_gap": 0.0,
436
+ "support_gap": 0.0,
437
+ "tau": 0.10322580228482307
438
+ },
439
+ "1375492786/train1": {
440
+ "base_success": 1.0,
441
+ "base_utility": 2.0,
442
+ "coverage": 0.0,
443
+ "execute_generated": 0.0,
444
+ "fallback_rate": 1.0,
445
+ "hidden_chart_oracle_success": 1.0,
446
+ "hidden_chart_oracle_utility": 2.0,
447
+ "outcome_ptr": 0.0,
448
+ "predicted_margin": -0.3663997567816525,
449
+ "proposal_oracle_success": 0.0,
450
+ "proposal_oracle_utility": 0.08251869678497314,
451
+ "selected_candidate_index": 2.0,
452
+ "selected_success": 1.0,
453
+ "selected_success_gain_over_base": 0.0,
454
+ "selected_utility": 2.0,
455
+ "selected_utility_gain_over_base": 0.0,
456
+ "selector_regret": 0.0,
457
+ "success_selector_gap": 0.0,
458
+ "success_support_gap": 1.0,
459
+ "support_gap": 1.9174813032150269,
460
+ "tau": 0.10322580228482307
461
+ },
462
+ "1375492786/train2": {
463
+ "base_success": 1.0,
464
+ "base_utility": 2.0,
465
+ "coverage": 0.0,
466
+ "execute_generated": 0.0,
467
+ "fallback_rate": 1.0,
468
+ "hidden_chart_oracle_success": 1.0,
469
+ "hidden_chart_oracle_utility": 2.0,
470
+ "outcome_ptr": 0.0,
471
+ "predicted_margin": -0.2028314963042361,
472
+ "proposal_oracle_success": 1.0,
473
+ "proposal_oracle_utility": 2.0,
474
+ "selected_candidate_index": 7.0,
475
+ "selected_success": 1.0,
476
+ "selected_success_gain_over_base": 0.0,
477
+ "selected_utility": 2.0,
478
+ "selected_utility_gain_over_base": 0.0,
479
+ "selector_regret": 0.0,
480
+ "success_selector_gap": 0.0,
481
+ "success_support_gap": 0.0,
482
+ "support_gap": 0.0,
483
+ "tau": 0.10322580228482307
484
+ },
485
+ "141984615/train0": {
486
+ "base_success": 0.0,
487
+ "base_utility": 0.10283424705266953,
488
+ "coverage": 0.0,
489
+ "execute_generated": 0.0,
490
+ "fallback_rate": 1.0,
491
+ "hidden_chart_oracle_success": 0.0,
492
+ "hidden_chart_oracle_utility": 0.2857772707939148,
493
+ "outcome_ptr": 0.0,
494
+ "predicted_margin": 0.0112655809567659,
495
+ "proposal_oracle_success": 0.0,
496
+ "proposal_oracle_utility": 0.007291515823453665,
497
+ "selected_candidate_index": 3.0,
498
+ "selected_success": 0.0,
499
+ "selected_success_gain_over_base": 0.0,
500
+ "selected_utility": 0.10283424705266953,
501
+ "selected_utility_gain_over_base": 0.0,
502
+ "selector_regret": 0.0,
503
+ "success_selector_gap": 0.0,
504
+ "success_support_gap": 0.0,
505
+ "support_gap": 0.27848575497046113,
506
+ "tau": 0.10322580228482307
507
+ },
508
+ "141984615/train1": {
509
+ "base_success": 0.0,
510
+ "base_utility": 0.10283424705266953,
511
+ "coverage": 1.0,
512
+ "execute_generated": 1.0,
513
+ "fallback_rate": 0.0,
514
+ "hidden_chart_oracle_success": 0.0,
515
+ "hidden_chart_oracle_utility": 0.2857772707939148,
516
+ "outcome_ptr": 0.0,
517
+ "predicted_margin": 0.10739113015380432,
518
+ "proposal_oracle_success": 0.0,
519
+ "proposal_oracle_utility": 0.037011127918958664,
520
+ "selected_candidate_index": 1.0,
521
+ "selected_success": 0.0,
522
+ "selected_success_gain_over_base": 0.0,
523
+ "selected_utility": 0.027254482731223106,
524
+ "selected_utility_gain_over_base": -0.07557976432144642,
525
+ "selector_regret": 0.009756645187735558,
526
+ "success_selector_gap": 0.0,
527
+ "success_support_gap": 0.0,
528
+ "support_gap": 0.24876614287495613,
529
+ "tau": 0.10322580228482307
530
+ },
531
+ "141984615/train2": {
532
+ "base_success": 0.0,
533
+ "base_utility": 0.10283424705266953,
534
+ "coverage": 0.0,
535
+ "execute_generated": 0.0,
536
+ "fallback_rate": 1.0,
537
+ "hidden_chart_oracle_success": 0.0,
538
+ "hidden_chart_oracle_utility": 0.2857772707939148,
539
+ "outcome_ptr": 0.0,
540
+ "predicted_margin": -0.12278991163863787,
541
+ "proposal_oracle_success": 0.0,
542
+ "proposal_oracle_utility": 0.030228236690163612,
543
+ "selected_candidate_index": 4.0,
544
+ "selected_success": 0.0,
545
+ "selected_success_gain_over_base": 0.0,
546
+ "selected_utility": 0.10283424705266953,
547
+ "selected_utility_gain_over_base": 0.0,
548
+ "selector_regret": 0.0,
549
+ "success_selector_gap": 0.0,
550
+ "success_support_gap": 0.0,
551
+ "support_gap": 0.2555490341037512,
552
+ "tau": 0.10322580228482307
553
+ },
554
+ "1461069713/train0": {
555
+ "base_success": 0.0,
556
+ "base_utility": 0.0939030647277832,
557
+ "coverage": 0.0,
558
+ "execute_generated": 0.0,
559
+ "fallback_rate": 1.0,
560
+ "hidden_chart_oracle_success": 0.0,
561
+ "hidden_chart_oracle_utility": 0.1593424528837204,
562
+ "outcome_ptr": 0.0,
563
+ "predicted_margin": 0.05558832163693243,
564
+ "proposal_oracle_success": 0.0,
565
+ "proposal_oracle_utility": 0.06019836664199829,
566
+ "selected_candidate_index": 3.0,
567
+ "selected_success": 0.0,
568
+ "selected_success_gain_over_base": 0.0,
569
+ "selected_utility": 0.0939030647277832,
570
+ "selected_utility_gain_over_base": 0.0,
571
+ "selector_regret": 0.0,
572
+ "success_selector_gap": 0.0,
573
+ "success_support_gap": 0.0,
574
+ "support_gap": 0.0991440862417221,
575
+ "tau": 0.10322580228482307
576
+ },
577
+ "1461069713/train1": {
578
+ "base_success": 0.0,
579
+ "base_utility": 0.0939030647277832,
580
+ "coverage": 1.0,
581
+ "execute_generated": 1.0,
582
+ "fallback_rate": 0.0,
583
+ "hidden_chart_oracle_success": 0.0,
584
+ "hidden_chart_oracle_utility": 0.1593424528837204,
585
+ "outcome_ptr": 0.0,
586
+ "predicted_margin": 0.23522842899622207,
587
+ "proposal_oracle_success": 0.0,
588
+ "proposal_oracle_utility": 0.016585012897849083,
589
+ "selected_candidate_index": 1.0,
590
+ "selected_success": 0.0,
591
+ "selected_success_gain_over_base": 0.0,
592
+ "selected_utility": 0.0066074528731405735,
593
+ "selected_utility_gain_over_base": -0.08729561185464263,
594
+ "selector_regret": 0.00997756002470851,
595
+ "success_selector_gap": 0.0,
596
+ "success_support_gap": 0.0,
597
+ "support_gap": 0.14275743998587132,
598
+ "tau": 0.10322580228482307
599
+ },
600
+ "1461069713/train2": {
601
+ "base_success": 0.0,
602
+ "base_utility": 0.0939030647277832,
603
+ "coverage": 1.0,
604
+ "execute_generated": 1.0,
605
+ "fallback_rate": 0.0,
606
+ "hidden_chart_oracle_success": 0.0,
607
+ "hidden_chart_oracle_utility": 0.1593424528837204,
608
+ "outcome_ptr": 1.0,
609
+ "predicted_margin": 0.12039683020844315,
610
+ "proposal_oracle_success": 0.0,
611
+ "proposal_oracle_utility": 0.09970339387655258,
612
+ "selected_candidate_index": 1.0,
613
+ "selected_success": 0.0,
614
+ "selected_success_gain_over_base": 0.0,
615
+ "selected_utility": 0.09168759733438492,
616
+ "selected_utility_gain_over_base": -0.002215467393398285,
617
+ "selector_regret": 0.008015796542167664,
618
+ "success_selector_gap": 0.0,
619
+ "success_support_gap": 0.0,
620
+ "support_gap": 0.059639059007167816,
621
+ "tau": 0.10322580228482307
622
+ },
623
+ "1480856810/train0": {
624
+ "base_success": 0.0,
625
+ "base_utility": 0.08703407645225525,
626
+ "coverage": 1.0,
627
+ "execute_generated": 1.0,
628
+ "fallback_rate": 0.0,
629
+ "hidden_chart_oracle_success": 0.0,
630
+ "hidden_chart_oracle_utility": 0.20948036015033722,
631
+ "outcome_ptr": 1.0,
632
+ "predicted_margin": 0.11448305520833224,
633
+ "proposal_oracle_success": 1.0,
634
+ "proposal_oracle_utility": 2.0,
635
+ "selected_candidate_index": 0.0,
636
+ "selected_success": 1.0,
637
+ "selected_success_gain_over_base": 1.0,
638
+ "selected_utility": 2.0,
639
+ "selected_utility_gain_over_base": 1.9129659235477448,
640
+ "selector_regret": 0.0,
641
+ "success_selector_gap": 0.0,
642
+ "success_support_gap": 0.0,
643
+ "support_gap": 0.0,
644
+ "tau": 0.10322580228482307
645
+ },
646
+ "1480856810/train1": {
647
+ "base_success": 0.0,
648
+ "base_utility": 0.08703407645225525,
649
+ "coverage": 1.0,
650
+ "execute_generated": 1.0,
651
+ "fallback_rate": 0.0,
652
+ "hidden_chart_oracle_success": 0.0,
653
+ "hidden_chart_oracle_utility": 0.20948036015033722,
654
+ "outcome_ptr": 1.0,
655
+ "predicted_margin": 0.34159067709064733,
656
+ "proposal_oracle_success": 1.0,
657
+ "proposal_oracle_utility": 2.0,
658
+ "selected_candidate_index": 6.0,
659
+ "selected_success": 1.0,
660
+ "selected_success_gain_over_base": 1.0,
661
+ "selected_utility": 2.0,
662
+ "selected_utility_gain_over_base": 1.9129659235477448,
663
+ "selector_regret": 0.0,
664
+ "success_selector_gap": 0.0,
665
+ "success_support_gap": 0.0,
666
+ "support_gap": 0.0,
667
+ "tau": 0.10322580228482307
668
+ },
669
+ "1480856810/train2": {
670
+ "base_success": 0.0,
671
+ "base_utility": 0.08703407645225525,
672
+ "coverage": 1.0,
673
+ "execute_generated": 1.0,
674
+ "fallback_rate": 0.0,
675
+ "hidden_chart_oracle_success": 0.0,
676
+ "hidden_chart_oracle_utility": 0.20948036015033722,
677
+ "outcome_ptr": 1.0,
678
+ "predicted_margin": 0.1290560623730016,
679
+ "proposal_oracle_success": 1.0,
680
+ "proposal_oracle_utility": 2.0,
681
+ "selected_candidate_index": 2.0,
682
+ "selected_success": 0.0,
683
+ "selected_success_gain_over_base": 0.0,
684
+ "selected_utility": 0.21894000470638275,
685
+ "selected_utility_gain_over_base": 0.1319059282541275,
686
+ "selector_regret": 1.7810599952936172,
687
+ "success_selector_gap": 1.0,
688
+ "success_support_gap": 0.0,
689
+ "support_gap": 0.0,
690
+ "tau": 0.10322580228482307
691
+ },
692
+ "1509980664/train0": {
693
+ "base_success": 0.0,
694
+ "base_utility": 0.8334100246429443,
695
+ "coverage": 0.0,
696
+ "execute_generated": 0.0,
697
+ "fallback_rate": 1.0,
698
+ "hidden_chart_oracle_success": 1.0,
699
+ "hidden_chart_oracle_utility": 2.0,
700
+ "outcome_ptr": 1.0,
701
+ "predicted_margin": 0.03262325878739534,
702
+ "proposal_oracle_success": 1.0,
703
+ "proposal_oracle_utility": 2.0,
704
+ "selected_candidate_index": 1.0,
705
+ "selected_success": 0.0,
706
+ "selected_success_gain_over_base": 0.0,
707
+ "selected_utility": 0.8334100246429443,
708
+ "selected_utility_gain_over_base": 0.0,
709
+ "selector_regret": 1.1665899753570557,
710
+ "success_selector_gap": 1.0,
711
+ "success_support_gap": 0.0,
712
+ "support_gap": 0.0,
713
+ "tau": 0.10322580228482307
714
+ },
715
+ "1509980664/train1": {
716
+ "base_success": 0.0,
717
+ "base_utility": 0.8334100246429443,
718
+ "coverage": 0.0,
719
+ "execute_generated": 0.0,
720
+ "fallback_rate": 1.0,
721
+ "hidden_chart_oracle_success": 1.0,
722
+ "hidden_chart_oracle_utility": 2.0,
723
+ "outcome_ptr": 1.0,
724
+ "predicted_margin": 0.05675633806543154,
725
+ "proposal_oracle_success": 1.0,
726
+ "proposal_oracle_utility": 2.0,
727
+ "selected_candidate_index": 7.0,
728
+ "selected_success": 0.0,
729
+ "selected_success_gain_over_base": 0.0,
730
+ "selected_utility": 0.8334100246429443,
731
+ "selected_utility_gain_over_base": 0.0,
732
+ "selector_regret": 1.1665899753570557,
733
+ "success_selector_gap": 1.0,
734
+ "success_support_gap": 0.0,
735
+ "support_gap": 0.0,
736
+ "tau": 0.10322580228482307
737
+ },
738
+ "1509980664/train2": {
739
+ "base_success": 0.0,
740
+ "base_utility": 0.8334100246429443,
741
+ "coverage": 0.0,
742
+ "execute_generated": 0.0,
743
+ "fallback_rate": 1.0,
744
+ "hidden_chart_oracle_success": 1.0,
745
+ "hidden_chart_oracle_utility": 2.0,
746
+ "outcome_ptr": 0.0,
747
+ "predicted_margin": 0.09855078704628988,
748
+ "proposal_oracle_success": 0.0,
749
+ "proposal_oracle_utility": 0.8162217140197754,
750
+ "selected_candidate_index": 3.0,
751
+ "selected_success": 0.0,
752
+ "selected_success_gain_over_base": 0.0,
753
+ "selected_utility": 0.8334100246429443,
754
+ "selected_utility_gain_over_base": 0.0,
755
+ "selector_regret": 0.0,
756
+ "success_selector_gap": 0.0,
757
+ "success_support_gap": 1.0,
758
+ "support_gap": 1.1837782859802246,
759
+ "tau": 0.10322580228482307
760
+ },
761
+ "1522272686/train0": {
762
+ "base_success": 0.0,
763
+ "base_utility": 0.5839591026306152,
764
+ "coverage": 0.0,
765
+ "execute_generated": 0.0,
766
+ "fallback_rate": 1.0,
767
+ "hidden_chart_oracle_success": 1.0,
768
+ "hidden_chart_oracle_utility": 2.0,
769
+ "outcome_ptr": 0.0,
770
+ "predicted_margin": -0.24386806647939827,
771
+ "proposal_oracle_success": 0.0,
772
+ "proposal_oracle_utility": 0.18325337767601013,
773
+ "selected_candidate_index": 4.0,
774
+ "selected_success": 0.0,
775
+ "selected_success_gain_over_base": 0.0,
776
+ "selected_utility": 0.5839591026306152,
777
+ "selected_utility_gain_over_base": 0.0,
778
+ "selector_regret": 0.0,
779
+ "success_selector_gap": 0.0,
780
+ "success_support_gap": 1.0,
781
+ "support_gap": 1.8167466223239899,
782
+ "tau": 0.10322580228482307
783
+ },
784
+ "1522272686/train1": {
785
+ "base_success": 0.0,
786
+ "base_utility": 0.5839591026306152,
787
+ "coverage": 0.0,
788
+ "execute_generated": 0.0,
789
+ "fallback_rate": 1.0,
790
+ "hidden_chart_oracle_success": 1.0,
791
+ "hidden_chart_oracle_utility": 2.0,
792
+ "outcome_ptr": 0.0,
793
+ "predicted_margin": -0.14675087965801709,
794
+ "proposal_oracle_success": 0.0,
795
+ "proposal_oracle_utility": 0.07014383375644684,
796
+ "selected_candidate_index": 4.0,
797
+ "selected_success": 0.0,
798
+ "selected_success_gain_over_base": 0.0,
799
+ "selected_utility": 0.5839591026306152,
800
+ "selected_utility_gain_over_base": 0.0,
801
+ "selector_regret": 0.0,
802
+ "success_selector_gap": 0.0,
803
+ "success_support_gap": 1.0,
804
+ "support_gap": 1.9298561662435532,
805
+ "tau": 0.10322580228482307
806
+ },
807
+ "1522272686/train2": {
808
+ "base_success": 0.0,
809
+ "base_utility": 0.5839591026306152,
810
+ "coverage": 0.0,
811
+ "execute_generated": 0.0,
812
+ "fallback_rate": 1.0,
813
+ "hidden_chart_oracle_success": 1.0,
814
+ "hidden_chart_oracle_utility": 2.0,
815
+ "outcome_ptr": 0.0,
816
+ "predicted_margin": -0.03770651742924813,
817
+ "proposal_oracle_success": 0.0,
818
+ "proposal_oracle_utility": 0.14937306940555573,
819
+ "selected_candidate_index": 4.0,
820
+ "selected_success": 0.0,
821
+ "selected_success_gain_over_base": 0.0,
822
+ "selected_utility": 0.5839591026306152,
823
+ "selected_utility_gain_over_base": 0.0,
824
+ "selector_regret": 0.0,
825
+ "success_selector_gap": 0.0,
826
+ "success_support_gap": 1.0,
827
+ "support_gap": 1.8506269305944443,
828
+ "tau": 0.10322580228482307
829
+ },
830
+ "1549324805/train0": {
831
+ "base_success": 0.0,
832
+ "base_utility": 0.10451483726501465,
833
+ "coverage": 0.0,
834
+ "execute_generated": 0.0,
835
+ "fallback_rate": 1.0,
836
+ "hidden_chart_oracle_success": 0.0,
837
+ "hidden_chart_oracle_utility": 0.2974560856819153,
838
+ "outcome_ptr": 0.0,
839
+ "predicted_margin": 0.07612026647778233,
840
+ "proposal_oracle_success": 0.0,
841
+ "proposal_oracle_utility": 0.0415392704308033,
842
+ "selected_candidate_index": 4.0,
843
+ "selected_success": 0.0,
844
+ "selected_success_gain_over_base": 0.0,
845
+ "selected_utility": 0.10451483726501465,
846
+ "selected_utility_gain_over_base": 0.0,
847
+ "selector_regret": 0.0,
848
+ "success_selector_gap": 0.0,
849
+ "success_support_gap": 0.0,
850
+ "support_gap": 0.255916815251112,
851
+ "tau": 0.10322580228482307
852
+ },
853
+ "1549324805/train1": {
854
+ "base_success": 0.0,
855
+ "base_utility": 0.10451483726501465,
856
+ "coverage": 0.0,
857
+ "execute_generated": 0.0,
858
+ "fallback_rate": 1.0,
859
+ "hidden_chart_oracle_success": 0.0,
860
+ "hidden_chart_oracle_utility": 0.2974560856819153,
861
+ "outcome_ptr": 1.0,
862
+ "predicted_margin": 0.037940358930246665,
863
+ "proposal_oracle_success": 0.0,
864
+ "proposal_oracle_utility": 0.19892151653766632,
865
+ "selected_candidate_index": 5.0,
866
+ "selected_success": 0.0,
867
+ "selected_success_gain_over_base": 0.0,
868
+ "selected_utility": 0.10451483726501465,
869
+ "selected_utility_gain_over_base": 0.0,
870
+ "selector_regret": 0.09440667927265167,
871
+ "success_selector_gap": 0.0,
872
+ "success_support_gap": 0.0,
873
+ "support_gap": 0.09853456914424896,
874
+ "tau": 0.10322580228482307
875
+ },
876
+ "1549324805/train2": {
877
+ "base_success": 0.0,
878
+ "base_utility": 0.10451483726501465,
879
+ "coverage": 1.0,
880
+ "execute_generated": 1.0,
881
+ "fallback_rate": 0.0,
882
+ "hidden_chart_oracle_success": 0.0,
883
+ "hidden_chart_oracle_utility": 0.2974560856819153,
884
+ "outcome_ptr": 0.0,
885
+ "predicted_margin": 0.2610975770137556,
886
+ "proposal_oracle_success": 0.0,
887
+ "proposal_oracle_utility": 0.09092829376459122,
888
+ "selected_candidate_index": 1.0,
889
+ "selected_success": 0.0,
890
+ "selected_success_gain_over_base": 0.0,
891
+ "selected_utility": 0.04690708592534065,
892
+ "selected_utility_gain_over_base": -0.057607751339673996,
893
+ "selector_regret": 0.044021207839250565,
894
+ "success_selector_gap": 0.0,
895
+ "success_support_gap": 0.0,
896
+ "support_gap": 0.20652779191732407,
897
+ "tau": 0.10322580228482307
898
+ },
899
+ "1578990937/train0": {
900
+ "base_success": 0.0,
901
+ "base_utility": 0.82413649559021,
902
+ "coverage": 0.0,
903
+ "execute_generated": 0.0,
904
+ "fallback_rate": 1.0,
905
+ "hidden_chart_oracle_success": 1.0,
906
+ "hidden_chart_oracle_utility": 2.0,
907
+ "outcome_ptr": 1.0,
908
+ "predicted_margin": 0.10318731059315213,
909
+ "proposal_oracle_success": 1.0,
910
+ "proposal_oracle_utility": 2.0,
911
+ "selected_candidate_index": 7.0,
912
+ "selected_success": 0.0,
913
+ "selected_success_gain_over_base": 0.0,
914
+ "selected_utility": 0.82413649559021,
915
+ "selected_utility_gain_over_base": 0.0,
916
+ "selector_regret": 1.17586350440979,
917
+ "success_selector_gap": 1.0,
918
+ "success_support_gap": 0.0,
919
+ "support_gap": 0.0,
920
+ "tau": 0.10322580228482307
921
+ },
922
+ "1578990937/train1": {
923
+ "base_success": 0.0,
924
+ "base_utility": 0.82413649559021,
925
+ "coverage": 1.0,
926
+ "execute_generated": 1.0,
927
+ "fallback_rate": 0.0,
928
+ "hidden_chart_oracle_success": 1.0,
929
+ "hidden_chart_oracle_utility": 2.0,
930
+ "outcome_ptr": 1.0,
931
+ "predicted_margin": 0.10918347583275737,
932
+ "proposal_oracle_success": 1.0,
933
+ "proposal_oracle_utility": 2.0,
934
+ "selected_candidate_index": 2.0,
935
+ "selected_success": 0.0,
936
+ "selected_success_gain_over_base": 0.0,
937
+ "selected_utility": 0.1305437982082367,
938
+ "selected_utility_gain_over_base": -0.6935926973819733,
939
+ "selector_regret": 1.8694562017917633,
940
+ "success_selector_gap": 1.0,
941
+ "success_support_gap": 0.0,
942
+ "support_gap": 0.0,
943
+ "tau": 0.10322580228482307
944
+ },
945
+ "1578990937/train2": {
946
+ "base_success": 0.0,
947
+ "base_utility": 0.82413649559021,
948
+ "coverage": 1.0,
949
+ "execute_generated": 1.0,
950
+ "fallback_rate": 0.0,
951
+ "hidden_chart_oracle_success": 1.0,
952
+ "hidden_chart_oracle_utility": 2.0,
953
+ "outcome_ptr": 1.0,
954
+ "predicted_margin": 0.1389029645691428,
955
+ "proposal_oracle_success": 1.0,
956
+ "proposal_oracle_utility": 2.0,
957
+ "selected_candidate_index": 2.0,
958
+ "selected_success": 1.0,
959
+ "selected_success_gain_over_base": 1.0,
960
+ "selected_utility": 2.0,
961
+ "selected_utility_gain_over_base": 1.17586350440979,
962
+ "selector_regret": 0.0,
963
+ "success_selector_gap": 0.0,
964
+ "success_support_gap": 0.0,
965
+ "support_gap": 0.0,
966
+ "tau": 0.10322580228482307
967
+ },
968
+ "1747178030/train0": {
969
+ "base_success": 1.0,
970
+ "base_utility": 2.0,
971
+ "coverage": 0.0,
972
+ "execute_generated": 0.0,
973
+ "fallback_rate": 1.0,
974
+ "hidden_chart_oracle_success": 1.0,
975
+ "hidden_chart_oracle_utility": 2.0,
976
+ "outcome_ptr": 0.0,
977
+ "predicted_margin": 0.0795986875833129,
978
+ "proposal_oracle_success": 1.0,
979
+ "proposal_oracle_utility": 2.0,
980
+ "selected_candidate_index": 1.0,
981
+ "selected_success": 1.0,
982
+ "selected_success_gain_over_base": 0.0,
983
+ "selected_utility": 2.0,
984
+ "selected_utility_gain_over_base": 0.0,
985
+ "selector_regret": 0.0,
986
+ "success_selector_gap": 0.0,
987
+ "success_support_gap": 0.0,
988
+ "support_gap": 0.0,
989
+ "tau": 0.10322580228482307
990
+ },
991
+ "1747178030/train1": {
992
+ "base_success": 1.0,
993
+ "base_utility": 2.0,
994
+ "coverage": 1.0,
995
+ "execute_generated": 1.0,
996
+ "fallback_rate": 0.0,
997
+ "hidden_chart_oracle_success": 1.0,
998
+ "hidden_chart_oracle_utility": 2.0,
999
+ "outcome_ptr": 0.0,
1000
+ "predicted_margin": 0.1350050049776791,
1001
+ "proposal_oracle_success": 1.0,
1002
+ "proposal_oracle_utility": 2.0,
1003
+ "selected_candidate_index": 4.0,
1004
+ "selected_success": 1.0,
1005
+ "selected_success_gain_over_base": 0.0,
1006
+ "selected_utility": 2.0,
1007
+ "selected_utility_gain_over_base": 0.0,
1008
+ "selector_regret": 0.0,
1009
+ "success_selector_gap": 0.0,
1010
+ "success_support_gap": 0.0,
1011
+ "support_gap": 0.0,
1012
+ "tau": 0.10322580228482307
1013
+ },
1014
+ "1747178030/train2": {
1015
+ "base_success": 1.0,
1016
+ "base_utility": 2.0,
1017
+ "coverage": 0.0,
1018
+ "execute_generated": 0.0,
1019
+ "fallback_rate": 1.0,
1020
+ "hidden_chart_oracle_success": 1.0,
1021
+ "hidden_chart_oracle_utility": 2.0,
1022
+ "outcome_ptr": 0.0,
1023
+ "predicted_margin": 0.09701624242728624,
1024
+ "proposal_oracle_success": 1.0,
1025
+ "proposal_oracle_utility": 2.0,
1026
+ "selected_candidate_index": 6.0,
1027
+ "selected_success": 1.0,
1028
+ "selected_success_gain_over_base": 0.0,
1029
+ "selected_utility": 2.0,
1030
+ "selected_utility_gain_over_base": 0.0,
1031
+ "selector_regret": 0.0,
1032
+ "success_selector_gap": 0.0,
1033
+ "success_support_gap": 0.0,
1034
+ "support_gap": 0.0,
1035
+ "tau": 0.10322580228482307
1036
+ },
1037
+ "1790188787/train0": {
1038
+ "base_success": 0.0,
1039
+ "base_utility": 0.12476746737957001,
1040
+ "coverage": 0.0,
1041
+ "execute_generated": 0.0,
1042
+ "fallback_rate": 1.0,
1043
+ "hidden_chart_oracle_success": 1.0,
1044
+ "hidden_chart_oracle_utility": 2.0,
1045
+ "outcome_ptr": 0.0,
1046
+ "predicted_margin": 0.09114133675253047,
1047
+ "proposal_oracle_success": 0.0,
1048
+ "proposal_oracle_utility": 0.09791924059391022,
1049
+ "selected_candidate_index": 1.0,
1050
+ "selected_success": 0.0,
1051
+ "selected_success_gain_over_base": 0.0,
1052
+ "selected_utility": 0.12476746737957001,
1053
+ "selected_utility_gain_over_base": 0.0,
1054
+ "selector_regret": 0.0,
1055
+ "success_selector_gap": 0.0,
1056
+ "success_support_gap": 1.0,
1057
+ "support_gap": 1.9020807594060898,
1058
+ "tau": 0.10322580228482307
1059
+ },
1060
+ "1790188787/train1": {
1061
+ "base_success": 0.0,
1062
+ "base_utility": 0.12476746737957001,
1063
+ "coverage": 0.0,
1064
+ "execute_generated": 0.0,
1065
+ "fallback_rate": 1.0,
1066
+ "hidden_chart_oracle_success": 1.0,
1067
+ "hidden_chart_oracle_utility": 2.0,
1068
+ "outcome_ptr": 1.0,
1069
+ "predicted_margin": 0.03217767457567711,
1070
+ "proposal_oracle_success": 0.0,
1071
+ "proposal_oracle_utility": 0.13031470775604248,
1072
+ "selected_candidate_index": 7.0,
1073
+ "selected_success": 0.0,
1074
+ "selected_success_gain_over_base": 0.0,
1075
+ "selected_utility": 0.12476746737957001,
1076
+ "selected_utility_gain_over_base": 0.0,
1077
+ "selector_regret": 0.005547240376472473,
1078
+ "success_selector_gap": 0.0,
1079
+ "success_support_gap": 1.0,
1080
+ "support_gap": 1.8696852922439575,
1081
+ "tau": 0.10322580228482307
1082
+ },
1083
+ "1790188787/train2": {
1084
+ "base_success": 0.0,
1085
+ "base_utility": 0.12476746737957001,
1086
+ "coverage": 0.0,
1087
+ "execute_generated": 0.0,
1088
+ "fallback_rate": 1.0,
1089
+ "hidden_chart_oracle_success": 1.0,
1090
+ "hidden_chart_oracle_utility": 2.0,
1091
+ "outcome_ptr": 1.0,
1092
+ "predicted_margin": 0.09038966104578511,
1093
+ "proposal_oracle_success": 1.0,
1094
+ "proposal_oracle_utility": 2.0,
1095
+ "selected_candidate_index": 1.0,
1096
+ "selected_success": 0.0,
1097
+ "selected_success_gain_over_base": 0.0,
1098
+ "selected_utility": 0.12476746737957001,
1099
+ "selected_utility_gain_over_base": 0.0,
1100
+ "selector_regret": 1.87523253262043,
1101
+ "success_selector_gap": 1.0,
1102
+ "success_support_gap": 0.0,
1103
+ "support_gap": 0.0,
1104
+ "tau": 0.10322580228482307
1105
+ },
1106
+ "1822210471/train0": {
1107
+ "base_success": 0.0,
1108
+ "base_utility": 0.07191530615091324,
1109
+ "coverage": 0.0,
1110
+ "execute_generated": 0.0,
1111
+ "fallback_rate": 1.0,
1112
+ "hidden_chart_oracle_success": 1.0,
1113
+ "hidden_chart_oracle_utility": 2.0,
1114
+ "outcome_ptr": 1.0,
1115
+ "predicted_margin": 0.029660176327400478,
1116
+ "proposal_oracle_success": 1.0,
1117
+ "proposal_oracle_utility": 2.0,
1118
+ "selected_candidate_index": 3.0,
1119
+ "selected_success": 0.0,
1120
+ "selected_success_gain_over_base": 0.0,
1121
+ "selected_utility": 0.07191530615091324,
1122
+ "selected_utility_gain_over_base": 0.0,
1123
+ "selector_regret": 1.9280846938490868,
1124
+ "success_selector_gap": 1.0,
1125
+ "success_support_gap": 0.0,
1126
+ "support_gap": 0.0,
1127
+ "tau": 0.10322580228482307
1128
+ },
1129
+ "1822210471/train1": {
1130
+ "base_success": 0.0,
1131
+ "base_utility": 0.07191530615091324,
1132
+ "coverage": 0.0,
1133
+ "execute_generated": 0.0,
1134
+ "fallback_rate": 1.0,
1135
+ "hidden_chart_oracle_success": 1.0,
1136
+ "hidden_chart_oracle_utility": 2.0,
1137
+ "outcome_ptr": 1.0,
1138
+ "predicted_margin": 0.05638544497648976,
1139
+ "proposal_oracle_success": 1.0,
1140
+ "proposal_oracle_utility": 2.0,
1141
+ "selected_candidate_index": 3.0,
1142
+ "selected_success": 0.0,
1143
+ "selected_success_gain_over_base": 0.0,
1144
+ "selected_utility": 0.07191530615091324,
1145
+ "selected_utility_gain_over_base": 0.0,
1146
+ "selector_regret": 1.9280846938490868,
1147
+ "success_selector_gap": 1.0,
1148
+ "success_support_gap": 0.0,
1149
+ "support_gap": 0.0,
1150
+ "tau": 0.10322580228482307
1151
+ },
1152
+ "1822210471/train2": {
1153
+ "base_success": 0.0,
1154
+ "base_utility": 0.07191530615091324,
1155
+ "coverage": 0.0,
1156
+ "execute_generated": 0.0,
1157
+ "fallback_rate": 1.0,
1158
+ "hidden_chart_oracle_success": 1.0,
1159
+ "hidden_chart_oracle_utility": 2.0,
1160
+ "outcome_ptr": 1.0,
1161
+ "predicted_margin": 0.04022648540053714,
1162
+ "proposal_oracle_success": 1.0,
1163
+ "proposal_oracle_utility": 2.0,
1164
+ "selected_candidate_index": 7.0,
1165
+ "selected_success": 0.0,
1166
+ "selected_success_gain_over_base": 0.0,
1167
+ "selected_utility": 0.07191530615091324,
1168
+ "selected_utility_gain_over_base": 0.0,
1169
+ "selector_regret": 1.9280846938490868,
1170
+ "success_selector_gap": 1.0,
1171
+ "success_support_gap": 0.0,
1172
+ "support_gap": 0.0,
1173
+ "tau": 0.10322580228482307
1174
+ },
1175
+ "1845489013/train0": {
1176
+ "base_success": 1.0,
1177
+ "base_utility": 2.0,
1178
+ "coverage": 0.0,
1179
+ "execute_generated": 0.0,
1180
+ "fallback_rate": 1.0,
1181
+ "hidden_chart_oracle_success": 1.0,
1182
+ "hidden_chart_oracle_utility": 2.0,
1183
+ "outcome_ptr": 0.0,
1184
+ "predicted_margin": 0.07542612499164425,
1185
+ "proposal_oracle_success": 1.0,
1186
+ "proposal_oracle_utility": 2.0,
1187
+ "selected_candidate_index": 5.0,
1188
+ "selected_success": 1.0,
1189
+ "selected_success_gain_over_base": 0.0,
1190
+ "selected_utility": 2.0,
1191
+ "selected_utility_gain_over_base": 0.0,
1192
+ "selector_regret": 0.0,
1193
+ "success_selector_gap": 0.0,
1194
+ "success_support_gap": 0.0,
1195
+ "support_gap": 0.0,
1196
+ "tau": 0.10322580228482307
1197
+ },
1198
+ "1845489013/train1": {
1199
+ "base_success": 1.0,
1200
+ "base_utility": 2.0,
1201
+ "coverage": 0.0,
1202
+ "execute_generated": 0.0,
1203
+ "fallback_rate": 1.0,
1204
+ "hidden_chart_oracle_success": 1.0,
1205
+ "hidden_chart_oracle_utility": 2.0,
1206
+ "outcome_ptr": 0.0,
1207
+ "predicted_margin": 0.07036421684892999,
1208
+ "proposal_oracle_success": 1.0,
1209
+ "proposal_oracle_utility": 2.0,
1210
+ "selected_candidate_index": 3.0,
1211
+ "selected_success": 1.0,
1212
+ "selected_success_gain_over_base": 0.0,
1213
+ "selected_utility": 2.0,
1214
+ "selected_utility_gain_over_base": 0.0,
1215
+ "selector_regret": 0.0,
1216
+ "success_selector_gap": 0.0,
1217
+ "success_support_gap": 0.0,
1218
+ "support_gap": 0.0,
1219
+ "tau": 0.10322580228482307
1220
+ },
1221
+ "1845489013/train2": {
1222
+ "base_success": 1.0,
1223
+ "base_utility": 2.0,
1224
+ "coverage": 0.0,
1225
+ "execute_generated": 0.0,
1226
+ "fallback_rate": 1.0,
1227
+ "hidden_chart_oracle_success": 1.0,
1228
+ "hidden_chart_oracle_utility": 2.0,
1229
+ "outcome_ptr": 0.0,
1230
+ "predicted_margin": 0.07957603786045438,
1231
+ "proposal_oracle_success": 1.0,
1232
+ "proposal_oracle_utility": 2.0,
1233
+ "selected_candidate_index": 6.0,
1234
+ "selected_success": 1.0,
1235
+ "selected_success_gain_over_base": 0.0,
1236
+ "selected_utility": 2.0,
1237
+ "selected_utility_gain_over_base": 0.0,
1238
+ "selector_regret": 0.0,
1239
+ "success_selector_gap": 0.0,
1240
+ "success_support_gap": 0.0,
1241
+ "support_gap": 0.0,
1242
+ "tau": 0.10322580228482307
1243
+ },
1244
+ "1999022574/train0": {
1245
+ "base_success": 1.0,
1246
+ "base_utility": 2.0,
1247
+ "coverage": 0.0,
1248
+ "execute_generated": 0.0,
1249
+ "fallback_rate": 1.0,
1250
+ "hidden_chart_oracle_success": 1.0,
1251
+ "hidden_chart_oracle_utility": 2.0,
1252
+ "outcome_ptr": 0.0,
1253
+ "predicted_margin": 0.02162026624278557,
1254
+ "proposal_oracle_success": 1.0,
1255
+ "proposal_oracle_utility": 2.0,
1256
+ "selected_candidate_index": 2.0,
1257
+ "selected_success": 1.0,
1258
+ "selected_success_gain_over_base": 0.0,
1259
+ "selected_utility": 2.0,
1260
+ "selected_utility_gain_over_base": 0.0,
1261
+ "selector_regret": 0.0,
1262
+ "success_selector_gap": 0.0,
1263
+ "success_support_gap": 0.0,
1264
+ "support_gap": 0.0,
1265
+ "tau": 0.10322580228482307
1266
+ },
1267
+ "1999022574/train1": {
1268
+ "base_success": 1.0,
1269
+ "base_utility": 2.0,
1270
+ "coverage": 0.0,
1271
+ "execute_generated": 0.0,
1272
+ "fallback_rate": 1.0,
1273
+ "hidden_chart_oracle_success": 1.0,
1274
+ "hidden_chart_oracle_utility": 2.0,
1275
+ "outcome_ptr": 0.0,
1276
+ "predicted_margin": 0.05584444198935358,
1277
+ "proposal_oracle_success": 1.0,
1278
+ "proposal_oracle_utility": 2.0,
1279
+ "selected_candidate_index": 4.0,
1280
+ "selected_success": 1.0,
1281
+ "selected_success_gain_over_base": 0.0,
1282
+ "selected_utility": 2.0,
1283
+ "selected_utility_gain_over_base": 0.0,
1284
+ "selector_regret": 0.0,
1285
+ "success_selector_gap": 0.0,
1286
+ "success_support_gap": 0.0,
1287
+ "support_gap": 0.0,
1288
+ "tau": 0.10322580228482307
1289
+ },
1290
+ "1999022574/train2": {
1291
+ "base_success": 1.0,
1292
+ "base_utility": 2.0,
1293
+ "coverage": 0.0,
1294
+ "execute_generated": 0.0,
1295
+ "fallback_rate": 1.0,
1296
+ "hidden_chart_oracle_success": 1.0,
1297
+ "hidden_chart_oracle_utility": 2.0,
1298
+ "outcome_ptr": 0.0,
1299
+ "predicted_margin": 0.060605829277106726,
1300
+ "proposal_oracle_success": 1.0,
1301
+ "proposal_oracle_utility": 2.0,
1302
+ "selected_candidate_index": 4.0,
1303
+ "selected_success": 1.0,
1304
+ "selected_success_gain_over_base": 0.0,
1305
+ "selected_utility": 2.0,
1306
+ "selected_utility_gain_over_base": 0.0,
1307
+ "selector_regret": 0.0,
1308
+ "success_selector_gap": 0.0,
1309
+ "success_support_gap": 0.0,
1310
+ "support_gap": 0.0,
1311
+ "tau": 0.10322580228482307
1312
+ },
1313
+ "2012211287/train0": {
1314
+ "base_success": 0.0,
1315
+ "base_utility": 0.7236053347587585,
1316
+ "coverage": 0.0,
1317
+ "execute_generated": 0.0,
1318
+ "fallback_rate": 1.0,
1319
+ "hidden_chart_oracle_success": 1.0,
1320
+ "hidden_chart_oracle_utility": 2.0,
1321
+ "outcome_ptr": 1.0,
1322
+ "predicted_margin": 0.07227334644188821,
1323
+ "proposal_oracle_success": 1.0,
1324
+ "proposal_oracle_utility": 2.0,
1325
+ "selected_candidate_index": 5.0,
1326
+ "selected_success": 0.0,
1327
+ "selected_success_gain_over_base": 0.0,
1328
+ "selected_utility": 0.7236053347587585,
1329
+ "selected_utility_gain_over_base": 0.0,
1330
+ "selector_regret": 1.2763946652412415,
1331
+ "success_selector_gap": 1.0,
1332
+ "success_support_gap": 0.0,
1333
+ "support_gap": 0.0,
1334
+ "tau": 0.10322580228482307
1335
+ },
1336
+ "2012211287/train1": {
1337
+ "base_success": 0.0,
1338
+ "base_utility": 0.7236053347587585,
1339
+ "coverage": 0.0,
1340
+ "execute_generated": 0.0,
1341
+ "fallback_rate": 1.0,
1342
+ "hidden_chart_oracle_success": 1.0,
1343
+ "hidden_chart_oracle_utility": 2.0,
1344
+ "outcome_ptr": 1.0,
1345
+ "predicted_margin": 0.008638727668705118,
1346
+ "proposal_oracle_success": 1.0,
1347
+ "proposal_oracle_utility": 2.0,
1348
+ "selected_candidate_index": 7.0,
1349
+ "selected_success": 0.0,
1350
+ "selected_success_gain_over_base": 0.0,
1351
+ "selected_utility": 0.7236053347587585,
1352
+ "selected_utility_gain_over_base": 0.0,
1353
+ "selector_regret": 1.2763946652412415,
1354
+ "success_selector_gap": 1.0,
1355
+ "success_support_gap": 0.0,
1356
+ "support_gap": 0.0,
1357
+ "tau": 0.10322580228482307
1358
+ },
1359
+ "2012211287/train2": {
1360
+ "base_success": 0.0,
1361
+ "base_utility": 0.7236053347587585,
1362
+ "coverage": 0.0,
1363
+ "execute_generated": 0.0,
1364
+ "fallback_rate": 1.0,
1365
+ "hidden_chart_oracle_success": 1.0,
1366
+ "hidden_chart_oracle_utility": 2.0,
1367
+ "outcome_ptr": 1.0,
1368
+ "predicted_margin": -0.020828488486956717,
1369
+ "proposal_oracle_success": 1.0,
1370
+ "proposal_oracle_utility": 2.0,
1371
+ "selected_candidate_index": 6.0,
1372
+ "selected_success": 0.0,
1373
+ "selected_success_gain_over_base": 0.0,
1374
+ "selected_utility": 0.7236053347587585,
1375
+ "selected_utility_gain_over_base": 0.0,
1376
+ "selector_regret": 1.2763946652412415,
1377
+ "success_selector_gap": 1.0,
1378
+ "success_support_gap": 0.0,
1379
+ "support_gap": 0.0,
1380
+ "tau": 0.10322580228482307
1381
+ },
1382
+ "2038594191/train0": {
1383
+ "base_success": 1.0,
1384
+ "base_utility": 2.0,
1385
+ "coverage": 0.0,
1386
+ "execute_generated": 0.0,
1387
+ "fallback_rate": 1.0,
1388
+ "hidden_chart_oracle_success": 1.0,
1389
+ "hidden_chart_oracle_utility": 2.0,
1390
+ "outcome_ptr": 0.0,
1391
+ "predicted_margin": -0.012327472859887986,
1392
+ "proposal_oracle_success": 1.0,
1393
+ "proposal_oracle_utility": 2.0,
1394
+ "selected_candidate_index": 3.0,
1395
+ "selected_success": 1.0,
1396
+ "selected_success_gain_over_base": 0.0,
1397
+ "selected_utility": 2.0,
1398
+ "selected_utility_gain_over_base": 0.0,
1399
+ "selector_regret": 0.0,
1400
+ "success_selector_gap": 0.0,
1401
+ "success_support_gap": 0.0,
1402
+ "support_gap": 0.0,
1403
+ "tau": 0.10322580228482307
1404
+ },
1405
+ "2038594191/train1": {
1406
+ "base_success": 1.0,
1407
+ "base_utility": 2.0,
1408
+ "coverage": 0.0,
1409
+ "execute_generated": 0.0,
1410
+ "fallback_rate": 1.0,
1411
+ "hidden_chart_oracle_success": 1.0,
1412
+ "hidden_chart_oracle_utility": 2.0,
1413
+ "outcome_ptr": 0.0,
1414
+ "predicted_margin": 0.0013949968247283923,
1415
+ "proposal_oracle_success": 1.0,
1416
+ "proposal_oracle_utility": 2.0,
1417
+ "selected_candidate_index": 0.0,
1418
+ "selected_success": 1.0,
1419
+ "selected_success_gain_over_base": 0.0,
1420
+ "selected_utility": 2.0,
1421
+ "selected_utility_gain_over_base": 0.0,
1422
+ "selector_regret": 0.0,
1423
+ "success_selector_gap": 0.0,
1424
+ "success_support_gap": 0.0,
1425
+ "support_gap": 0.0,
1426
+ "tau": 0.10322580228482307
1427
+ },
1428
+ "2038594191/train2": {
1429
+ "base_success": 1.0,
1430
+ "base_utility": 2.0,
1431
+ "coverage": 0.0,
1432
+ "execute_generated": 0.0,
1433
+ "fallback_rate": 1.0,
1434
+ "hidden_chart_oracle_success": 1.0,
1435
+ "hidden_chart_oracle_utility": 2.0,
1436
+ "outcome_ptr": 0.0,
1437
+ "predicted_margin": 0.001100725693189697,
1438
+ "proposal_oracle_success": 1.0,
1439
+ "proposal_oracle_utility": 2.0,
1440
+ "selected_candidate_index": 1.0,
1441
+ "selected_success": 1.0,
1442
+ "selected_success_gain_over_base": 0.0,
1443
+ "selected_utility": 2.0,
1444
+ "selected_utility_gain_over_base": 0.0,
1445
+ "selector_regret": 0.0,
1446
+ "success_selector_gap": 0.0,
1447
+ "success_support_gap": 0.0,
1448
+ "support_gap": 0.0,
1449
+ "tau": 0.10322580228482307
1450
+ },
1451
+ "2054899126/train0": {
1452
+ "base_success": 1.0,
1453
+ "base_utility": 2.0,
1454
+ "coverage": 0.5,
1455
+ "execute_generated": 0.5,
1456
+ "fallback_rate": 0.5,
1457
+ "hidden_chart_oracle_success": 1.0,
1458
+ "hidden_chart_oracle_utility": 2.0,
1459
+ "outcome_ptr": 0.0,
1460
+ "predicted_margin": 0.017351904963364015,
1461
+ "proposal_oracle_success": 0.5,
1462
+ "proposal_oracle_utility": 1.077283151447773,
1463
+ "selected_candidate_index": 3.0,
1464
+ "selected_success": 1.0,
1465
+ "selected_success_gain_over_base": 0.0,
1466
+ "selected_utility": 2.0,
1467
+ "selected_utility_gain_over_base": 0.0,
1468
+ "selector_regret": 0.0,
1469
+ "success_selector_gap": 0.0,
1470
+ "success_support_gap": 0.5,
1471
+ "support_gap": 0.922716848552227,
1472
+ "tau": 0.10322580228482307
1473
+ },
1474
+ "2054899126/train1": {
1475
+ "base_success": 1.0,
1476
+ "base_utility": 2.0,
1477
+ "coverage": 0.5,
1478
+ "execute_generated": 0.5,
1479
+ "fallback_rate": 0.5,
1480
+ "hidden_chart_oracle_success": 1.0,
1481
+ "hidden_chart_oracle_utility": 2.0,
1482
+ "outcome_ptr": 0.0,
1483
+ "predicted_margin": 0.06668682370964799,
1484
+ "proposal_oracle_success": 0.5,
1485
+ "proposal_oracle_utility": 1.0696808770298958,
1486
+ "selected_candidate_index": 3.0,
1487
+ "selected_success": 1.0,
1488
+ "selected_success_gain_over_base": 0.0,
1489
+ "selected_utility": 2.0,
1490
+ "selected_utility_gain_over_base": 0.0,
1491
+ "selector_regret": 0.0,
1492
+ "success_selector_gap": 0.0,
1493
+ "success_support_gap": 0.5,
1494
+ "support_gap": 0.9303191229701042,
1495
+ "tau": 0.10322580228482307
1496
+ },
1497
+ "2054899126/train2": {
1498
+ "base_success": 1.0,
1499
+ "base_utility": 2.0,
1500
+ "coverage": 0.0,
1501
+ "execute_generated": 0.0,
1502
+ "fallback_rate": 1.0,
1503
+ "hidden_chart_oracle_success": 1.0,
1504
+ "hidden_chart_oracle_utility": 2.0,
1505
+ "outcome_ptr": 0.0,
1506
+ "predicted_margin": -0.06439727728823853,
1507
+ "proposal_oracle_success": 0.5,
1508
+ "proposal_oracle_utility": 1.0455262660980225,
1509
+ "selected_candidate_index": 0.5,
1510
+ "selected_success": 1.0,
1511
+ "selected_success_gain_over_base": 0.0,
1512
+ "selected_utility": 2.0,
1513
+ "selected_utility_gain_over_base": 0.0,
1514
+ "selector_regret": 0.0,
1515
+ "success_selector_gap": 0.0,
1516
+ "success_support_gap": 0.5,
1517
+ "support_gap": 0.9544737339019775,
1518
+ "tau": 0.10322580228482307
1519
+ },
1520
+ "2061635888/train0": {
1521
+ "base_success": 0.0,
1522
+ "base_utility": 0.05225050449371338,
1523
+ "coverage": 0.0,
1524
+ "execute_generated": 0.0,
1525
+ "fallback_rate": 1.0,
1526
+ "hidden_chart_oracle_success": 1.0,
1527
+ "hidden_chart_oracle_utility": 2.0,
1528
+ "outcome_ptr": 1.0,
1529
+ "predicted_margin": -0.008988580649135547,
1530
+ "proposal_oracle_success": 0.0,
1531
+ "proposal_oracle_utility": 0.09799930453300476,
1532
+ "selected_candidate_index": 2.0,
1533
+ "selected_success": 0.0,
1534
+ "selected_success_gain_over_base": 0.0,
1535
+ "selected_utility": 0.05225050449371338,
1536
+ "selected_utility_gain_over_base": 0.0,
1537
+ "selector_regret": 0.04574880003929138,
1538
+ "success_selector_gap": 0.0,
1539
+ "success_support_gap": 1.0,
1540
+ "support_gap": 1.9020006954669952,
1541
+ "tau": 0.10322580228482307
1542
+ },
1543
+ "2061635888/train1": {
1544
+ "base_success": 0.0,
1545
+ "base_utility": 0.05225050449371338,
1546
+ "coverage": 0.0,
1547
+ "execute_generated": 0.0,
1548
+ "fallback_rate": 1.0,
1549
+ "hidden_chart_oracle_success": 1.0,
1550
+ "hidden_chart_oracle_utility": 2.0,
1551
+ "outcome_ptr": 1.0,
1552
+ "predicted_margin": 0.0391771378386657,
1553
+ "proposal_oracle_success": 0.0,
1554
+ "proposal_oracle_utility": 0.07580336928367615,
1555
+ "selected_candidate_index": 1.0,
1556
+ "selected_success": 0.0,
1557
+ "selected_success_gain_over_base": 0.0,
1558
+ "selected_utility": 0.05225050449371338,
1559
+ "selected_utility_gain_over_base": 0.0,
1560
+ "selector_regret": 0.02355286478996277,
1561
+ "success_selector_gap": 0.0,
1562
+ "success_support_gap": 1.0,
1563
+ "support_gap": 1.9241966307163239,
1564
+ "tau": 0.10322580228482307
1565
+ },
1566
+ "2061635888/train2": {
1567
+ "base_success": 0.0,
1568
+ "base_utility": 0.05225050449371338,
1569
+ "coverage": 0.0,
1570
+ "execute_generated": 0.0,
1571
+ "fallback_rate": 1.0,
1572
+ "hidden_chart_oracle_success": 1.0,
1573
+ "hidden_chart_oracle_utility": 2.0,
1574
+ "outcome_ptr": 1.0,
1575
+ "predicted_margin": 0.017664101020798448,
1576
+ "proposal_oracle_success": 0.0,
1577
+ "proposal_oracle_utility": 0.11105132102966309,
1578
+ "selected_candidate_index": 6.0,
1579
+ "selected_success": 0.0,
1580
+ "selected_success_gain_over_base": 0.0,
1581
+ "selected_utility": 0.05225050449371338,
1582
+ "selected_utility_gain_over_base": 0.0,
1583
+ "selector_regret": 0.05880081653594971,
1584
+ "success_selector_gap": 0.0,
1585
+ "success_support_gap": 1.0,
1586
+ "support_gap": 1.888948678970337,
1587
+ "tau": 0.10322580228482307
1588
+ },
1589
+ "211245842/train0": {
1590
+ "base_success": 0.0,
1591
+ "base_utility": 0.19090841710567474,
1592
+ "coverage": 0.0,
1593
+ "execute_generated": 0.0,
1594
+ "fallback_rate": 1.0,
1595
+ "hidden_chart_oracle_success": 1.0,
1596
+ "hidden_chart_oracle_utility": 2.0,
1597
+ "outcome_ptr": 1.0,
1598
+ "predicted_margin": 0.031065660093159524,
1599
+ "proposal_oracle_success": 1.0,
1600
+ "proposal_oracle_utility": 2.0,
1601
+ "selected_candidate_index": 3.0,
1602
+ "selected_success": 0.0,
1603
+ "selected_success_gain_over_base": 0.0,
1604
+ "selected_utility": 0.19090841710567474,
1605
+ "selected_utility_gain_over_base": 0.0,
1606
+ "selector_regret": 1.8090915828943253,
1607
+ "success_selector_gap": 1.0,
1608
+ "success_support_gap": 0.0,
1609
+ "support_gap": 0.0,
1610
+ "tau": 0.10322580228482307
1611
+ },
1612
+ "211245842/train1": {
1613
+ "base_success": 0.0,
1614
+ "base_utility": 0.19090841710567474,
1615
+ "coverage": 0.0,
1616
+ "execute_generated": 0.0,
1617
+ "fallback_rate": 1.0,
1618
+ "hidden_chart_oracle_success": 1.0,
1619
+ "hidden_chart_oracle_utility": 2.0,
1620
+ "outcome_ptr": 1.0,
1621
+ "predicted_margin": 0.09139080941968937,
1622
+ "proposal_oracle_success": 0.0,
1623
+ "proposal_oracle_utility": 0.21295717358589172,
1624
+ "selected_candidate_index": 6.0,
1625
+ "selected_success": 0.0,
1626
+ "selected_success_gain_over_base": 0.0,
1627
+ "selected_utility": 0.19090841710567474,
1628
+ "selected_utility_gain_over_base": 0.0,
1629
+ "selector_regret": 0.02204875648021698,
1630
+ "success_selector_gap": 0.0,
1631
+ "success_support_gap": 1.0,
1632
+ "support_gap": 1.7870428264141083,
1633
+ "tau": 0.10322580228482307
1634
+ },
1635
+ "211245842/train2": {
1636
+ "base_success": 0.0,
1637
+ "base_utility": 0.19090841710567474,
1638
+ "coverage": 0.0,
1639
+ "execute_generated": 0.0,
1640
+ "fallback_rate": 1.0,
1641
+ "hidden_chart_oracle_success": 1.0,
1642
+ "hidden_chart_oracle_utility": 2.0,
1643
+ "outcome_ptr": 1.0,
1644
+ "predicted_margin": 0.04780007356962993,
1645
+ "proposal_oracle_success": 0.0,
1646
+ "proposal_oracle_utility": 0.21523693203926086,
1647
+ "selected_candidate_index": 0.0,
1648
+ "selected_success": 0.0,
1649
+ "selected_success_gain_over_base": 0.0,
1650
+ "selected_utility": 0.19090841710567474,
1651
+ "selected_utility_gain_over_base": 0.0,
1652
+ "selector_regret": 0.02432851493358612,
1653
+ "success_selector_gap": 0.0,
1654
+ "success_support_gap": 1.0,
1655
+ "support_gap": 1.7847630679607391,
1656
+ "tau": 0.10322580228482307
1657
+ },
1658
+ "278008524/train0": {
1659
+ "base_success": 0.0,
1660
+ "base_utility": 0.042734989896416664,
1661
+ "coverage": 0.0,
1662
+ "execute_generated": 0.0,
1663
+ "fallback_rate": 1.0,
1664
+ "hidden_chart_oracle_success": 0.5,
1665
+ "hidden_chart_oracle_utility": 1.1139086410403252,
1666
+ "outcome_ptr": 1.0,
1667
+ "predicted_margin": -0.03661210287452511,
1668
+ "proposal_oracle_success": 0.5,
1669
+ "proposal_oracle_utility": 1.0287093725055456,
1670
+ "selected_candidate_index": 4.5,
1671
+ "selected_success": 0.0,
1672
+ "selected_success_gain_over_base": 0.0,
1673
+ "selected_utility": 0.042734989896416664,
1674
+ "selected_utility_gain_over_base": 0.0,
1675
+ "selector_regret": 0.985974382609129,
1676
+ "success_selector_gap": 0.5,
1677
+ "success_support_gap": 0.0,
1678
+ "support_gap": 0.08519926853477955,
1679
+ "tau": 0.10322580228482307
1680
+ },
1681
+ "278008524/train1": {
1682
+ "base_success": 0.0,
1683
+ "base_utility": 0.042734989896416664,
1684
+ "coverage": 0.5,
1685
+ "execute_generated": 0.5,
1686
+ "fallback_rate": 0.5,
1687
+ "hidden_chart_oracle_success": 0.5,
1688
+ "hidden_chart_oracle_utility": 1.1139086410403252,
1689
+ "outcome_ptr": 1.0,
1690
+ "predicted_margin": 0.04337092305047836,
1691
+ "proposal_oracle_success": 0.5,
1692
+ "proposal_oracle_utility": 1.0376745946705341,
1693
+ "selected_candidate_index": 3.5,
1694
+ "selected_success": 0.0,
1695
+ "selected_success_gain_over_base": 0.0,
1696
+ "selected_utility": 0.05399738810956478,
1697
+ "selected_utility_gain_over_base": 0.011262398213148117,
1698
+ "selector_regret": 0.9836772065609694,
1699
+ "success_selector_gap": 0.5,
1700
+ "success_support_gap": 0.0,
1701
+ "support_gap": 0.07623404636979103,
1702
+ "tau": 0.10322580228482307
1703
+ },
1704
+ "278008524/train2": {
1705
+ "base_success": 0.0,
1706
+ "base_utility": 0.042734989896416664,
1707
+ "coverage": 0.0,
1708
+ "execute_generated": 0.0,
1709
+ "fallback_rate": 1.0,
1710
+ "hidden_chart_oracle_success": 0.5,
1711
+ "hidden_chart_oracle_utility": 1.1139086410403252,
1712
+ "outcome_ptr": 1.0,
1713
+ "predicted_margin": 0.014342230361280552,
1714
+ "proposal_oracle_success": 0.5,
1715
+ "proposal_oracle_utility": 1.1359489113092422,
1716
+ "selected_candidate_index": 3.5,
1717
+ "selected_success": 0.0,
1718
+ "selected_success_gain_over_base": 0.0,
1719
+ "selected_utility": 0.042734989896416664,
1720
+ "selected_utility_gain_over_base": 0.0,
1721
+ "selector_regret": 1.0932139214128256,
1722
+ "success_selector_gap": 0.5,
1723
+ "success_support_gap": 0.0,
1724
+ "support_gap": 0.0,
1725
+ "tau": 0.10322580228482307
1726
+ },
1727
+ "307068879/train0": {
1728
+ "base_success": 0.0,
1729
+ "base_utility": 0.6000905632972717,
1730
+ "coverage": 1.0,
1731
+ "execute_generated": 1.0,
1732
+ "fallback_rate": 0.0,
1733
+ "hidden_chart_oracle_success": 1.0,
1734
+ "hidden_chart_oracle_utility": 2.0,
1735
+ "outcome_ptr": 0.0,
1736
+ "predicted_margin": 0.13593551823980266,
1737
+ "proposal_oracle_success": 0.0,
1738
+ "proposal_oracle_utility": 0.2115575522184372,
1739
+ "selected_candidate_index": 3.0,
1740
+ "selected_success": 0.0,
1741
+ "selected_success_gain_over_base": 0.0,
1742
+ "selected_utility": 0.20163558423519135,
1743
+ "selected_utility_gain_over_base": -0.3984549790620804,
1744
+ "selector_regret": 0.00992196798324585,
1745
+ "success_selector_gap": 0.0,
1746
+ "success_support_gap": 1.0,
1747
+ "support_gap": 1.7884424477815628,
1748
+ "tau": 0.10322580228482307
1749
+ },
1750
+ "307068879/train1": {
1751
+ "base_success": 0.0,
1752
+ "base_utility": 0.6000905632972717,
1753
+ "coverage": 0.0,
1754
+ "execute_generated": 0.0,
1755
+ "fallback_rate": 1.0,
1756
+ "hidden_chart_oracle_success": 1.0,
1757
+ "hidden_chart_oracle_utility": 2.0,
1758
+ "outcome_ptr": 0.0,
1759
+ "predicted_margin": 0.0010682845366857263,
1760
+ "proposal_oracle_success": 0.0,
1761
+ "proposal_oracle_utility": 0.5876073241233826,
1762
+ "selected_candidate_index": 5.0,
1763
+ "selected_success": 0.0,
1764
+ "selected_success_gain_over_base": 0.0,
1765
+ "selected_utility": 0.6000905632972717,
1766
+ "selected_utility_gain_over_base": 0.0,
1767
+ "selector_regret": 0.0,
1768
+ "success_selector_gap": 0.0,
1769
+ "success_support_gap": 1.0,
1770
+ "support_gap": 1.4123926758766174,
1771
+ "tau": 0.10322580228482307
1772
+ },
1773
+ "307068879/train2": {
1774
+ "base_success": 0.0,
1775
+ "base_utility": 0.6000905632972717,
1776
+ "coverage": 0.0,
1777
+ "execute_generated": 0.0,
1778
+ "fallback_rate": 1.0,
1779
+ "hidden_chart_oracle_success": 1.0,
1780
+ "hidden_chart_oracle_utility": 2.0,
1781
+ "outcome_ptr": 0.0,
1782
+ "predicted_margin": 0.03087667088250734,
1783
+ "proposal_oracle_success": 0.0,
1784
+ "proposal_oracle_utility": 0.22727864980697632,
1785
+ "selected_candidate_index": 5.0,
1786
+ "selected_success": 0.0,
1787
+ "selected_success_gain_over_base": 0.0,
1788
+ "selected_utility": 0.6000905632972717,
1789
+ "selected_utility_gain_over_base": 0.0,
1790
+ "selector_regret": 0.0,
1791
+ "success_selector_gap": 0.0,
1792
+ "success_support_gap": 1.0,
1793
+ "support_gap": 1.7727213501930237,
1794
+ "tau": 0.10322580228482307
1795
+ },
1796
+ "412102563/train0": {
1797
+ "base_success": 1.0,
1798
+ "base_utility": 2.0,
1799
+ "coverage": 0.0,
1800
+ "execute_generated": 0.0,
1801
+ "fallback_rate": 1.0,
1802
+ "hidden_chart_oracle_success": 1.0,
1803
+ "hidden_chart_oracle_utility": 2.0,
1804
+ "outcome_ptr": 0.0,
1805
+ "predicted_margin": 0.02404992196724428,
1806
+ "proposal_oracle_success": 1.0,
1807
+ "proposal_oracle_utility": 2.0,
1808
+ "selected_candidate_index": 2.0,
1809
+ "selected_success": 1.0,
1810
+ "selected_success_gain_over_base": 0.0,
1811
+ "selected_utility": 2.0,
1812
+ "selected_utility_gain_over_base": 0.0,
1813
+ "selector_regret": 0.0,
1814
+ "success_selector_gap": 0.0,
1815
+ "success_support_gap": 0.0,
1816
+ "support_gap": 0.0,
1817
+ "tau": 0.10322580228482307
1818
+ },
1819
+ "412102563/train1": {
1820
+ "base_success": 1.0,
1821
+ "base_utility": 2.0,
1822
+ "coverage": 0.0,
1823
+ "execute_generated": 0.0,
1824
+ "fallback_rate": 1.0,
1825
+ "hidden_chart_oracle_success": 1.0,
1826
+ "hidden_chart_oracle_utility": 2.0,
1827
+ "outcome_ptr": 0.0,
1828
+ "predicted_margin": 0.052528880139410684,
1829
+ "proposal_oracle_success": 1.0,
1830
+ "proposal_oracle_utility": 2.0,
1831
+ "selected_candidate_index": 7.0,
1832
+ "selected_success": 1.0,
1833
+ "selected_success_gain_over_base": 0.0,
1834
+ "selected_utility": 2.0,
1835
+ "selected_utility_gain_over_base": 0.0,
1836
+ "selector_regret": 0.0,
1837
+ "success_selector_gap": 0.0,
1838
+ "success_support_gap": 0.0,
1839
+ "support_gap": 0.0,
1840
+ "tau": 0.10322580228482307
1841
+ },
1842
+ "412102563/train2": {
1843
+ "base_success": 1.0,
1844
+ "base_utility": 2.0,
1845
+ "coverage": 0.0,
1846
+ "execute_generated": 0.0,
1847
+ "fallback_rate": 1.0,
1848
+ "hidden_chart_oracle_success": 1.0,
1849
+ "hidden_chart_oracle_utility": 2.0,
1850
+ "outcome_ptr": 0.0,
1851
+ "predicted_margin": 0.07441605133316857,
1852
+ "proposal_oracle_success": 1.0,
1853
+ "proposal_oracle_utility": 2.0,
1854
+ "selected_candidate_index": 4.0,
1855
+ "selected_success": 1.0,
1856
+ "selected_success_gain_over_base": 0.0,
1857
+ "selected_utility": 2.0,
1858
+ "selected_utility_gain_over_base": 0.0,
1859
+ "selector_regret": 0.0,
1860
+ "success_selector_gap": 0.0,
1861
+ "success_support_gap": 0.0,
1862
+ "support_gap": 0.0,
1863
+ "tau": 0.10322580228482307
1864
+ },
1865
+ "440577711/train0": {
1866
+ "base_success": 0.0,
1867
+ "base_utility": 0.036152780055999756,
1868
+ "coverage": 1.0,
1869
+ "execute_generated": 1.0,
1870
+ "fallback_rate": 0.0,
1871
+ "hidden_chart_oracle_success": 1.0,
1872
+ "hidden_chart_oracle_utility": 2.0,
1873
+ "outcome_ptr": 1.0,
1874
+ "predicted_margin": 0.18180403057646763,
1875
+ "proposal_oracle_success": 1.0,
1876
+ "proposal_oracle_utility": 2.0,
1877
+ "selected_candidate_index": 1.0,
1878
+ "selected_success": 1.0,
1879
+ "selected_success_gain_over_base": 1.0,
1880
+ "selected_utility": 2.0,
1881
+ "selected_utility_gain_over_base": 1.9638472199440002,
1882
+ "selector_regret": 0.0,
1883
+ "success_selector_gap": 0.0,
1884
+ "success_support_gap": 0.0,
1885
+ "support_gap": 0.0,
1886
+ "tau": 0.10322580228482307
1887
+ },
1888
+ "440577711/train1": {
1889
+ "base_success": 0.0,
1890
+ "base_utility": 0.036152780055999756,
1891
+ "coverage": 1.0,
1892
+ "execute_generated": 1.0,
1893
+ "fallback_rate": 0.0,
1894
+ "hidden_chart_oracle_success": 1.0,
1895
+ "hidden_chart_oracle_utility": 2.0,
1896
+ "outcome_ptr": 1.0,
1897
+ "predicted_margin": 0.314855596033475,
1898
+ "proposal_oracle_success": 1.0,
1899
+ "proposal_oracle_utility": 2.0,
1900
+ "selected_candidate_index": 6.0,
1901
+ "selected_success": 1.0,
1902
+ "selected_success_gain_over_base": 1.0,
1903
+ "selected_utility": 2.0,
1904
+ "selected_utility_gain_over_base": 1.9638472199440002,
1905
+ "selector_regret": 0.0,
1906
+ "success_selector_gap": 0.0,
1907
+ "success_support_gap": 0.0,
1908
+ "support_gap": 0.0,
1909
+ "tau": 0.10322580228482307
1910
+ },
1911
+ "440577711/train2": {
1912
+ "base_success": 0.0,
1913
+ "base_utility": 0.036152780055999756,
1914
+ "coverage": 1.0,
1915
+ "execute_generated": 1.0,
1916
+ "fallback_rate": 0.0,
1917
+ "hidden_chart_oracle_success": 1.0,
1918
+ "hidden_chart_oracle_utility": 2.0,
1919
+ "outcome_ptr": 1.0,
1920
+ "predicted_margin": 0.1664010416335164,
1921
+ "proposal_oracle_success": 1.0,
1922
+ "proposal_oracle_utility": 2.0,
1923
+ "selected_candidate_index": 1.0,
1924
+ "selected_success": 1.0,
1925
+ "selected_success_gain_over_base": 1.0,
1926
+ "selected_utility": 2.0,
1927
+ "selected_utility_gain_over_base": 1.9638472199440002,
1928
+ "selector_regret": 0.0,
1929
+ "success_selector_gap": 0.0,
1930
+ "success_support_gap": 0.0,
1931
+ "support_gap": 0.0,
1932
+ "tau": 0.10322580228482307
1933
+ },
1934
+ "441046872/train0": {
1935
+ "base_success": 0.0,
1936
+ "base_utility": 0.6999551653862,
1937
+ "coverage": 0.0,
1938
+ "execute_generated": 0.0,
1939
+ "fallback_rate": 1.0,
1940
+ "hidden_chart_oracle_success": 1.0,
1941
+ "hidden_chart_oracle_utility": 2.0,
1942
+ "outcome_ptr": 0.0,
1943
+ "predicted_margin": 0.08693488284592116,
1944
+ "proposal_oracle_success": 0.0,
1945
+ "proposal_oracle_utility": 0.5148025155067444,
1946
+ "selected_candidate_index": 0.0,
1947
+ "selected_success": 0.0,
1948
+ "selected_success_gain_over_base": 0.0,
1949
+ "selected_utility": 0.6999551653862,
1950
+ "selected_utility_gain_over_base": 0.0,
1951
+ "selector_regret": 0.0,
1952
+ "success_selector_gap": 0.0,
1953
+ "success_support_gap": 1.0,
1954
+ "support_gap": 1.4851974844932556,
1955
+ "tau": 0.10322580228482307
1956
+ },
1957
+ "441046872/train1": {
1958
+ "base_success": 0.0,
1959
+ "base_utility": 0.6999551653862,
1960
+ "coverage": 0.0,
1961
+ "execute_generated": 0.0,
1962
+ "fallback_rate": 1.0,
1963
+ "hidden_chart_oracle_success": 1.0,
1964
+ "hidden_chart_oracle_utility": 2.0,
1965
+ "outcome_ptr": 1.0,
1966
+ "predicted_margin": 0.006713192112029656,
1967
+ "proposal_oracle_success": 0.0,
1968
+ "proposal_oracle_utility": 0.7233588695526123,
1969
+ "selected_candidate_index": 2.0,
1970
+ "selected_success": 0.0,
1971
+ "selected_success_gain_over_base": 0.0,
1972
+ "selected_utility": 0.6999551653862,
1973
+ "selected_utility_gain_over_base": 0.0,
1974
+ "selector_regret": 0.023403704166412354,
1975
+ "success_selector_gap": 0.0,
1976
+ "success_support_gap": 1.0,
1977
+ "support_gap": 1.2766411304473877,
1978
+ "tau": 0.10322580228482307
1979
+ },
1980
+ "441046872/train2": {
1981
+ "base_success": 0.0,
1982
+ "base_utility": 0.6999551653862,
1983
+ "coverage": 0.0,
1984
+ "execute_generated": 0.0,
1985
+ "fallback_rate": 1.0,
1986
+ "hidden_chart_oracle_success": 1.0,
1987
+ "hidden_chart_oracle_utility": 2.0,
1988
+ "outcome_ptr": 1.0,
1989
+ "predicted_margin": 0.06487491969714901,
1990
+ "proposal_oracle_success": 0.0,
1991
+ "proposal_oracle_utility": 0.7130289077758789,
1992
+ "selected_candidate_index": 1.0,
1993
+ "selected_success": 0.0,
1994
+ "selected_success_gain_over_base": 0.0,
1995
+ "selected_utility": 0.6999551653862,
1996
+ "selected_utility_gain_over_base": 0.0,
1997
+ "selector_regret": 0.013073742389678955,
1998
+ "success_selector_gap": 0.0,
1999
+ "success_support_gap": 1.0,
2000
+ "support_gap": 1.286971092224121,
2001
+ "tau": 0.10322580228482307
2002
+ },
2003
+ "509430392/train0": {
2004
+ "base_success": 0.0,
2005
+ "base_utility": 2.2351741790771484e-07,
2006
+ "coverage": 0.0,
2007
+ "execute_generated": 0.0,
2008
+ "fallback_rate": 1.0,
2009
+ "hidden_chart_oracle_success": 1.0,
2010
+ "hidden_chart_oracle_utility": 2.0,
2011
+ "outcome_ptr": 1.0,
2012
+ "predicted_margin": 0.08918827453095743,
2013
+ "proposal_oracle_success": 1.0,
2014
+ "proposal_oracle_utility": 2.0,
2015
+ "selected_candidate_index": 4.0,
2016
+ "selected_success": 0.0,
2017
+ "selected_success_gain_over_base": 0.0,
2018
+ "selected_utility": 2.2351741790771484e-07,
2019
+ "selected_utility_gain_over_base": 0.0,
2020
+ "selector_regret": 1.999999776482582,
2021
+ "success_selector_gap": 1.0,
2022
+ "success_support_gap": 0.0,
2023
+ "support_gap": 0.0,
2024
+ "tau": 0.10322580228482307
2025
+ },
2026
+ "509430392/train1": {
2027
+ "base_success": 0.0,
2028
+ "base_utility": 2.2351741790771484e-07,
2029
+ "coverage": 1.0,
2030
+ "execute_generated": 1.0,
2031
+ "fallback_rate": 0.0,
2032
+ "hidden_chart_oracle_success": 1.0,
2033
+ "hidden_chart_oracle_utility": 2.0,
2034
+ "outcome_ptr": 1.0,
2035
+ "predicted_margin": 0.13410303028042397,
2036
+ "proposal_oracle_success": 0.0,
2037
+ "proposal_oracle_utility": 6.164610385894775e-05,
2038
+ "selected_candidate_index": 2.0,
2039
+ "selected_success": 0.0,
2040
+ "selected_success_gain_over_base": 0.0,
2041
+ "selected_utility": 1.519918441772461e-06,
2042
+ "selected_utility_gain_over_base": 1.296401023864746e-06,
2043
+ "selector_regret": 6.012618541717529e-05,
2044
+ "success_selector_gap": 0.0,
2045
+ "success_support_gap": 1.0,
2046
+ "support_gap": 1.999938353896141,
2047
+ "tau": 0.10322580228482307
2048
+ },
2049
+ "509430392/train2": {
2050
+ "base_success": 0.0,
2051
+ "base_utility": 2.2351741790771484e-07,
2052
+ "coverage": 0.0,
2053
+ "execute_generated": 0.0,
2054
+ "fallback_rate": 1.0,
2055
+ "hidden_chart_oracle_success": 1.0,
2056
+ "hidden_chart_oracle_utility": 2.0,
2057
+ "outcome_ptr": 1.0,
2058
+ "predicted_margin": 0.009310188915263855,
2059
+ "proposal_oracle_success": 0.0,
2060
+ "proposal_oracle_utility": 5.424022674560547e-06,
2061
+ "selected_candidate_index": 5.0,
2062
+ "selected_success": 0.0,
2063
+ "selected_success_gain_over_base": 0.0,
2064
+ "selected_utility": 2.2351741790771484e-07,
2065
+ "selected_utility_gain_over_base": 0.0,
2066
+ "selector_regret": 5.200505256652832e-06,
2067
+ "success_selector_gap": 0.0,
2068
+ "success_support_gap": 1.0,
2069
+ "support_gap": 1.9999945759773254,
2070
+ "tau": 0.10322580228482307
2071
+ },
2072
+ "585904130/train0": {
2073
+ "base_success": 0.0,
2074
+ "base_utility": 0.18171554803848267,
2075
+ "coverage": 0.0,
2076
+ "execute_generated": 0.0,
2077
+ "fallback_rate": 1.0,
2078
+ "hidden_chart_oracle_success": 1.0,
2079
+ "hidden_chart_oracle_utility": 2.0,
2080
+ "outcome_ptr": 1.0,
2081
+ "predicted_margin": -0.1003597485778477,
2082
+ "proposal_oracle_success": 1.0,
2083
+ "proposal_oracle_utility": 2.0,
2084
+ "selected_candidate_index": 0.0,
2085
+ "selected_success": 0.0,
2086
+ "selected_success_gain_over_base": 0.0,
2087
+ "selected_utility": 0.18171554803848267,
2088
+ "selected_utility_gain_over_base": 0.0,
2089
+ "selector_regret": 1.8182844519615173,
2090
+ "success_selector_gap": 1.0,
2091
+ "success_support_gap": 0.0,
2092
+ "support_gap": 0.0,
2093
+ "tau": 0.10322580228482307
2094
+ },
2095
+ "585904130/train1": {
2096
+ "base_success": 0.0,
2097
+ "base_utility": 0.18171554803848267,
2098
+ "coverage": 0.0,
2099
+ "execute_generated": 0.0,
2100
+ "fallback_rate": 1.0,
2101
+ "hidden_chart_oracle_success": 1.0,
2102
+ "hidden_chart_oracle_utility": 2.0,
2103
+ "outcome_ptr": 1.0,
2104
+ "predicted_margin": -0.06992988938957599,
2105
+ "proposal_oracle_success": 1.0,
2106
+ "proposal_oracle_utility": 2.0,
2107
+ "selected_candidate_index": 0.0,
2108
+ "selected_success": 0.0,
2109
+ "selected_success_gain_over_base": 0.0,
2110
+ "selected_utility": 0.18171554803848267,
2111
+ "selected_utility_gain_over_base": 0.0,
2112
+ "selector_regret": 1.8182844519615173,
2113
+ "success_selector_gap": 1.0,
2114
+ "success_support_gap": 0.0,
2115
+ "support_gap": 0.0,
2116
+ "tau": 0.10322580228482307
2117
+ },
2118
+ "585904130/train2": {
2119
+ "base_success": 0.0,
2120
+ "base_utility": 0.18171554803848267,
2121
+ "coverage": 0.0,
2122
+ "execute_generated": 0.0,
2123
+ "fallback_rate": 1.0,
2124
+ "hidden_chart_oracle_success": 1.0,
2125
+ "hidden_chart_oracle_utility": 2.0,
2126
+ "outcome_ptr": 1.0,
2127
+ "predicted_margin": -0.08372863408577144,
2128
+ "proposal_oracle_success": 1.0,
2129
+ "proposal_oracle_utility": 2.0,
2130
+ "selected_candidate_index": 0.0,
2131
+ "selected_success": 0.0,
2132
+ "selected_success_gain_over_base": 0.0,
2133
+ "selected_utility": 0.18171554803848267,
2134
+ "selected_utility_gain_over_base": 0.0,
2135
+ "selector_regret": 1.8182844519615173,
2136
+ "success_selector_gap": 1.0,
2137
+ "success_support_gap": 0.0,
2138
+ "support_gap": 0.0,
2139
+ "tau": 0.10322580228482307
2140
+ },
2141
+ "608016290/train0": {
2142
+ "base_success": 0.0,
2143
+ "base_utility": 0.725701630115509,
2144
+ "coverage": 0.0,
2145
+ "execute_generated": 0.0,
2146
+ "fallback_rate": 1.0,
2147
+ "hidden_chart_oracle_success": 1.0,
2148
+ "hidden_chart_oracle_utility": 2.0,
2149
+ "outcome_ptr": 1.0,
2150
+ "predicted_margin": 0.09419178502546363,
2151
+ "proposal_oracle_success": 1.0,
2152
+ "proposal_oracle_utility": 2.0,
2153
+ "selected_candidate_index": 3.0,
2154
+ "selected_success": 0.0,
2155
+ "selected_success_gain_over_base": 0.0,
2156
+ "selected_utility": 0.725701630115509,
2157
+ "selected_utility_gain_over_base": 0.0,
2158
+ "selector_regret": 1.274298369884491,
2159
+ "success_selector_gap": 1.0,
2160
+ "success_support_gap": 0.0,
2161
+ "support_gap": 0.0,
2162
+ "tau": 0.10322580228482307
2163
+ },
2164
+ "608016290/train1": {
2165
+ "base_success": 0.0,
2166
+ "base_utility": 0.725701630115509,
2167
+ "coverage": 0.0,
2168
+ "execute_generated": 0.0,
2169
+ "fallback_rate": 1.0,
2170
+ "hidden_chart_oracle_success": 1.0,
2171
+ "hidden_chart_oracle_utility": 2.0,
2172
+ "outcome_ptr": 1.0,
2173
+ "predicted_margin": 0.06378367779574469,
2174
+ "proposal_oracle_success": 1.0,
2175
+ "proposal_oracle_utility": 2.0,
2176
+ "selected_candidate_index": 1.0,
2177
+ "selected_success": 0.0,
2178
+ "selected_success_gain_over_base": 0.0,
2179
+ "selected_utility": 0.725701630115509,
2180
+ "selected_utility_gain_over_base": 0.0,
2181
+ "selector_regret": 1.274298369884491,
2182
+ "success_selector_gap": 1.0,
2183
+ "success_support_gap": 0.0,
2184
+ "support_gap": 0.0,
2185
+ "tau": 0.10322580228482307
2186
+ },
2187
+ "608016290/train2": {
2188
+ "base_success": 0.0,
2189
+ "base_utility": 0.725701630115509,
2190
+ "coverage": 0.0,
2191
+ "execute_generated": 0.0,
2192
+ "fallback_rate": 1.0,
2193
+ "hidden_chart_oracle_success": 1.0,
2194
+ "hidden_chart_oracle_utility": 2.0,
2195
+ "outcome_ptr": 0.0,
2196
+ "predicted_margin": 0.09752490294074964,
2197
+ "proposal_oracle_success": 0.0,
2198
+ "proposal_oracle_utility": 0.6825780272483826,
2199
+ "selected_candidate_index": 2.0,
2200
+ "selected_success": 0.0,
2201
+ "selected_success_gain_over_base": 0.0,
2202
+ "selected_utility": 0.725701630115509,
2203
+ "selected_utility_gain_over_base": 0.0,
2204
+ "selector_regret": 0.0,
2205
+ "success_selector_gap": 0.0,
2206
+ "success_support_gap": 1.0,
2207
+ "support_gap": 1.3174219727516174,
2208
+ "tau": 0.10322580228482307
2209
+ },
2210
+ "626486813/train0": {
2211
+ "base_success": 0.0,
2212
+ "base_utility": 0.00509834272088483,
2213
+ "coverage": 1.0,
2214
+ "execute_generated": 1.0,
2215
+ "fallback_rate": 0.0,
2216
+ "hidden_chart_oracle_success": 0.0,
2217
+ "hidden_chart_oracle_utility": 0.16558238118886948,
2218
+ "outcome_ptr": 1.0,
2219
+ "predicted_margin": 0.11969968119788658,
2220
+ "proposal_oracle_success": 0.0,
2221
+ "proposal_oracle_utility": 0.043898362666368484,
2222
+ "selected_candidate_index": 5.0,
2223
+ "selected_success": 0.0,
2224
+ "selected_success_gain_over_base": 0.0,
2225
+ "selected_utility": 0.009900043485686183,
2226
+ "selected_utility_gain_over_base": 0.004801700764801353,
2227
+ "selector_regret": 0.0339983191806823,
2228
+ "success_selector_gap": 0.0,
2229
+ "success_support_gap": 0.0,
2230
+ "support_gap": 0.12241578102111816,
2231
+ "tau": 0.10322580228482307
2232
+ },
2233
+ "626486813/train1": {
2234
+ "base_success": 0.0,
2235
+ "base_utility": 0.00509834272088483,
2236
+ "coverage": 1.0,
2237
+ "execute_generated": 1.0,
2238
+ "fallback_rate": 0.0,
2239
+ "hidden_chart_oracle_success": 0.0,
2240
+ "hidden_chart_oracle_utility": 0.16558238118886948,
2241
+ "outcome_ptr": 1.0,
2242
+ "predicted_margin": 0.16641122274358117,
2243
+ "proposal_oracle_success": 0.0,
2244
+ "proposal_oracle_utility": 0.030555277597159147,
2245
+ "selected_candidate_index": 4.0,
2246
+ "selected_success": 0.0,
2247
+ "selected_success_gain_over_base": 0.0,
2248
+ "selected_utility": 0.019377360120415688,
2249
+ "selected_utility_gain_over_base": 0.014279017399530858,
2250
+ "selector_regret": 0.01117791747674346,
2251
+ "success_selector_gap": 0.0,
2252
+ "success_support_gap": 0.0,
2253
+ "support_gap": 0.13502710359171033,
2254
+ "tau": 0.10322580228482307
2255
+ },
2256
+ "626486813/train2": {
2257
+ "base_success": 0.0,
2258
+ "base_utility": 0.00509834272088483,
2259
+ "coverage": 0.0,
2260
+ "execute_generated": 0.0,
2261
+ "fallback_rate": 1.0,
2262
+ "hidden_chart_oracle_success": 0.0,
2263
+ "hidden_chart_oracle_utility": 0.16558238118886948,
2264
+ "outcome_ptr": 1.0,
2265
+ "predicted_margin": 0.048332577437914415,
2266
+ "proposal_oracle_success": 0.0,
2267
+ "proposal_oracle_utility": 0.03856666944921017,
2268
+ "selected_candidate_index": 1.0,
2269
+ "selected_success": 0.0,
2270
+ "selected_success_gain_over_base": 0.0,
2271
+ "selected_utility": 0.00509834272088483,
2272
+ "selected_utility_gain_over_base": 0.0,
2273
+ "selector_regret": 0.03346832672832534,
2274
+ "success_selector_gap": 0.0,
2275
+ "success_support_gap": 0.0,
2276
+ "support_gap": 0.1270157117396593,
2277
+ "tau": 0.10322580228482307
2278
+ },
2279
+ "636680137/train0": {
2280
+ "base_success": 0.0,
2281
+ "base_utility": 0.08313504606485367,
2282
+ "coverage": 1.0,
2283
+ "execute_generated": 1.0,
2284
+ "fallback_rate": 0.0,
2285
+ "hidden_chart_oracle_success": 0.0,
2286
+ "hidden_chart_oracle_utility": 0.17056432366371155,
2287
+ "outcome_ptr": 1.0,
2288
+ "predicted_margin": 0.1672723810719819,
2289
+ "proposal_oracle_success": 0.0,
2290
+ "proposal_oracle_utility": 0.08826130628585815,
2291
+ "selected_candidate_index": 1.0,
2292
+ "selected_success": 0.0,
2293
+ "selected_success_gain_over_base": 0.0,
2294
+ "selected_utility": 0.08826130628585815,
2295
+ "selected_utility_gain_over_base": 0.005126260221004486,
2296
+ "selector_regret": 0.0,
2297
+ "success_selector_gap": 0.0,
2298
+ "success_support_gap": 0.0,
2299
+ "support_gap": 0.0823030173778534,
2300
+ "tau": 0.10322580228482307
2301
+ },
2302
+ "636680137/train1": {
2303
+ "base_success": 0.0,
2304
+ "base_utility": 0.08313504606485367,
2305
+ "coverage": 0.0,
2306
+ "execute_generated": 0.0,
2307
+ "fallback_rate": 1.0,
2308
+ "hidden_chart_oracle_success": 0.0,
2309
+ "hidden_chart_oracle_utility": 0.17056432366371155,
2310
+ "outcome_ptr": 0.0,
2311
+ "predicted_margin": 0.09650428584630603,
2312
+ "proposal_oracle_success": 0.0,
2313
+ "proposal_oracle_utility": 0.06741118431091309,
2314
+ "selected_candidate_index": 1.0,
2315
+ "selected_success": 0.0,
2316
+ "selected_success_gain_over_base": 0.0,
2317
+ "selected_utility": 0.08313504606485367,
2318
+ "selected_utility_gain_over_base": 0.0,
2319
+ "selector_regret": 0.0,
2320
+ "success_selector_gap": 0.0,
2321
+ "success_support_gap": 0.0,
2322
+ "support_gap": 0.10315313935279846,
2323
+ "tau": 0.10322580228482307
2324
+ },
2325
+ "636680137/train2": {
2326
+ "base_success": 0.0,
2327
+ "base_utility": 0.08313504606485367,
2328
+ "coverage": 0.0,
2329
+ "execute_generated": 0.0,
2330
+ "fallback_rate": 1.0,
2331
+ "hidden_chart_oracle_success": 0.0,
2332
+ "hidden_chart_oracle_utility": 0.17056432366371155,
2333
+ "outcome_ptr": 0.0,
2334
+ "predicted_margin": 0.0915647194899811,
2335
+ "proposal_oracle_success": 0.0,
2336
+ "proposal_oracle_utility": 0.0756516233086586,
2337
+ "selected_candidate_index": 7.0,
2338
+ "selected_success": 0.0,
2339
+ "selected_success_gain_over_base": 0.0,
2340
+ "selected_utility": 0.08313504606485367,
2341
+ "selected_utility_gain_over_base": 0.0,
2342
+ "selector_regret": 0.0,
2343
+ "success_selector_gap": 0.0,
2344
+ "success_support_gap": 0.0,
2345
+ "support_gap": 0.09491270035505295,
2346
+ "tau": 0.10322580228482307
2347
+ },
2348
+ "655884605/train0": {
2349
+ "base_success": 0.0,
2350
+ "base_utility": 0.16300667822360992,
2351
+ "coverage": 0.0,
2352
+ "execute_generated": 0.0,
2353
+ "fallback_rate": 1.0,
2354
+ "hidden_chart_oracle_success": 1.0,
2355
+ "hidden_chart_oracle_utility": 2.0,
2356
+ "outcome_ptr": 0.0,
2357
+ "predicted_margin": -0.028147324820558327,
2358
+ "proposal_oracle_success": 0.0,
2359
+ "proposal_oracle_utility": 0.13732868432998657,
2360
+ "selected_candidate_index": 6.0,
2361
+ "selected_success": 0.0,
2362
+ "selected_success_gain_over_base": 0.0,
2363
+ "selected_utility": 0.16300667822360992,
2364
+ "selected_utility_gain_over_base": 0.0,
2365
+ "selector_regret": 0.0,
2366
+ "success_selector_gap": 0.0,
2367
+ "success_support_gap": 1.0,
2368
+ "support_gap": 1.8626713156700134,
2369
+ "tau": 0.10322580228482307
2370
+ },
2371
+ "655884605/train1": {
2372
+ "base_success": 0.0,
2373
+ "base_utility": 0.16300667822360992,
2374
+ "coverage": 0.0,
2375
+ "execute_generated": 0.0,
2376
+ "fallback_rate": 1.0,
2377
+ "hidden_chart_oracle_success": 1.0,
2378
+ "hidden_chart_oracle_utility": 2.0,
2379
+ "outcome_ptr": 1.0,
2380
+ "predicted_margin": -0.024645828875293,
2381
+ "proposal_oracle_success": 0.0,
2382
+ "proposal_oracle_utility": 0.1651531159877777,
2383
+ "selected_candidate_index": 7.0,
2384
+ "selected_success": 0.0,
2385
+ "selected_success_gain_over_base": 0.0,
2386
+ "selected_utility": 0.16300667822360992,
2387
+ "selected_utility_gain_over_base": 0.0,
2388
+ "selector_regret": 0.0021464377641677856,
2389
+ "success_selector_gap": 0.0,
2390
+ "success_support_gap": 1.0,
2391
+ "support_gap": 1.8348468840122223,
2392
+ "tau": 0.10322580228482307
2393
+ },
2394
+ "655884605/train2": {
2395
+ "base_success": 0.0,
2396
+ "base_utility": 0.16300667822360992,
2397
+ "coverage": 0.0,
2398
+ "execute_generated": 0.0,
2399
+ "fallback_rate": 1.0,
2400
+ "hidden_chart_oracle_success": 1.0,
2401
+ "hidden_chart_oracle_utility": 2.0,
2402
+ "outcome_ptr": 1.0,
2403
+ "predicted_margin": 0.027389502211879177,
2404
+ "proposal_oracle_success": 0.0,
2405
+ "proposal_oracle_utility": 0.21447330713272095,
2406
+ "selected_candidate_index": 7.0,
2407
+ "selected_success": 0.0,
2408
+ "selected_success_gain_over_base": 0.0,
2409
+ "selected_utility": 0.16300667822360992,
2410
+ "selected_utility_gain_over_base": 0.0,
2411
+ "selector_regret": 0.05146662890911102,
2412
+ "success_selector_gap": 0.0,
2413
+ "success_support_gap": 1.0,
2414
+ "support_gap": 1.785526692867279,
2415
+ "tau": 0.10322580228482307
2416
+ },
2417
+ "672328262/train0": {
2418
+ "base_success": 1.0,
2419
+ "base_utility": 2.0,
2420
+ "coverage": 0.0,
2421
+ "execute_generated": 0.0,
2422
+ "fallback_rate": 1.0,
2423
+ "hidden_chart_oracle_success": 1.0,
2424
+ "hidden_chart_oracle_utility": 2.0,
2425
+ "outcome_ptr": 0.0,
2426
+ "predicted_margin": 0.014765937563623946,
2427
+ "proposal_oracle_success": 1.0,
2428
+ "proposal_oracle_utility": 2.0,
2429
+ "selected_candidate_index": 1.0,
2430
+ "selected_success": 1.0,
2431
+ "selected_success_gain_over_base": 0.0,
2432
+ "selected_utility": 2.0,
2433
+ "selected_utility_gain_over_base": 0.0,
2434
+ "selector_regret": 0.0,
2435
+ "success_selector_gap": 0.0,
2436
+ "success_support_gap": 0.0,
2437
+ "support_gap": 0.0,
2438
+ "tau": 0.10322580228482307
2439
+ },
2440
+ "672328262/train1": {
2441
+ "base_success": 1.0,
2442
+ "base_utility": 2.0,
2443
+ "coverage": 0.0,
2444
+ "execute_generated": 0.0,
2445
+ "fallback_rate": 1.0,
2446
+ "hidden_chart_oracle_success": 1.0,
2447
+ "hidden_chart_oracle_utility": 2.0,
2448
+ "outcome_ptr": 0.0,
2449
+ "predicted_margin": 0.01697110302863744,
2450
+ "proposal_oracle_success": 1.0,
2451
+ "proposal_oracle_utility": 2.0,
2452
+ "selected_candidate_index": 0.0,
2453
+ "selected_success": 1.0,
2454
+ "selected_success_gain_over_base": 0.0,
2455
+ "selected_utility": 2.0,
2456
+ "selected_utility_gain_over_base": 0.0,
2457
+ "selector_regret": 0.0,
2458
+ "success_selector_gap": 0.0,
2459
+ "success_support_gap": 0.0,
2460
+ "support_gap": 0.0,
2461
+ "tau": 0.10322580228482307
2462
+ },
2463
+ "672328262/train2": {
2464
+ "base_success": 1.0,
2465
+ "base_utility": 2.0,
2466
+ "coverage": 0.0,
2467
+ "execute_generated": 0.0,
2468
+ "fallback_rate": 1.0,
2469
+ "hidden_chart_oracle_success": 1.0,
2470
+ "hidden_chart_oracle_utility": 2.0,
2471
+ "outcome_ptr": 0.0,
2472
+ "predicted_margin": -0.016167574303126375,
2473
+ "proposal_oracle_success": 1.0,
2474
+ "proposal_oracle_utility": 2.0,
2475
+ "selected_candidate_index": 1.0,
2476
+ "selected_success": 1.0,
2477
+ "selected_success_gain_over_base": 0.0,
2478
+ "selected_utility": 2.0,
2479
+ "selected_utility_gain_over_base": 0.0,
2480
+ "selector_regret": 0.0,
2481
+ "success_selector_gap": 0.0,
2482
+ "success_support_gap": 0.0,
2483
+ "support_gap": 0.0,
2484
+ "tau": 0.10322580228482307
2485
+ },
2486
+ "701667486/train0": {
2487
+ "base_success": 1.0,
2488
+ "base_utility": 2.0,
2489
+ "coverage": 0.0,
2490
+ "execute_generated": 0.0,
2491
+ "fallback_rate": 1.0,
2492
+ "hidden_chart_oracle_success": 1.0,
2493
+ "hidden_chart_oracle_utility": 2.0,
2494
+ "outcome_ptr": 0.0,
2495
+ "predicted_margin": 0.04665577151979795,
2496
+ "proposal_oracle_success": 1.0,
2497
+ "proposal_oracle_utility": 2.0,
2498
+ "selected_candidate_index": 2.0,
2499
+ "selected_success": 1.0,
2500
+ "selected_success_gain_over_base": 0.0,
2501
+ "selected_utility": 2.0,
2502
+ "selected_utility_gain_over_base": 0.0,
2503
+ "selector_regret": 0.0,
2504
+ "success_selector_gap": 0.0,
2505
+ "success_support_gap": 0.0,
2506
+ "support_gap": 0.0,
2507
+ "tau": 0.10322580228482307
2508
+ },
2509
+ "701667486/train1": {
2510
+ "base_success": 1.0,
2511
+ "base_utility": 2.0,
2512
+ "coverage": 0.0,
2513
+ "execute_generated": 0.0,
2514
+ "fallback_rate": 1.0,
2515
+ "hidden_chart_oracle_success": 1.0,
2516
+ "hidden_chart_oracle_utility": 2.0,
2517
+ "outcome_ptr": 0.0,
2518
+ "predicted_margin": 0.06468789057317349,
2519
+ "proposal_oracle_success": 1.0,
2520
+ "proposal_oracle_utility": 2.0,
2521
+ "selected_candidate_index": 1.0,
2522
+ "selected_success": 1.0,
2523
+ "selected_success_gain_over_base": 0.0,
2524
+ "selected_utility": 2.0,
2525
+ "selected_utility_gain_over_base": 0.0,
2526
+ "selector_regret": 0.0,
2527
+ "success_selector_gap": 0.0,
2528
+ "success_support_gap": 0.0,
2529
+ "support_gap": 0.0,
2530
+ "tau": 0.10322580228482307
2531
+ },
2532
+ "701667486/train2": {
2533
+ "base_success": 1.0,
2534
+ "base_utility": 2.0,
2535
+ "coverage": 1.0,
2536
+ "execute_generated": 1.0,
2537
+ "fallback_rate": 0.0,
2538
+ "hidden_chart_oracle_success": 1.0,
2539
+ "hidden_chart_oracle_utility": 2.0,
2540
+ "outcome_ptr": 0.0,
2541
+ "predicted_margin": 0.1202657721119895,
2542
+ "proposal_oracle_success": 1.0,
2543
+ "proposal_oracle_utility": 2.0,
2544
+ "selected_candidate_index": 1.0,
2545
+ "selected_success": 1.0,
2546
+ "selected_success_gain_over_base": 0.0,
2547
+ "selected_utility": 2.0,
2548
+ "selected_utility_gain_over_base": 0.0,
2549
+ "selector_regret": 0.0,
2550
+ "success_selector_gap": 0.0,
2551
+ "success_support_gap": 0.0,
2552
+ "support_gap": 0.0,
2553
+ "tau": 0.10322580228482307
2554
+ },
2555
+ "747260437/train0": {
2556
+ "base_success": 1.0,
2557
+ "base_utility": 2.0,
2558
+ "coverage": 0.0,
2559
+ "execute_generated": 0.0,
2560
+ "fallback_rate": 1.0,
2561
+ "hidden_chart_oracle_success": 0.0,
2562
+ "hidden_chart_oracle_utility": 0.2786632478237152,
2563
+ "outcome_ptr": 0.0,
2564
+ "predicted_margin": -0.10220617671360782,
2565
+ "proposal_oracle_success": 0.0,
2566
+ "proposal_oracle_utility": 0.10839563608169556,
2567
+ "selected_candidate_index": 4.0,
2568
+ "selected_success": 1.0,
2569
+ "selected_success_gain_over_base": 0.0,
2570
+ "selected_utility": 2.0,
2571
+ "selected_utility_gain_over_base": 0.0,
2572
+ "selector_regret": 0.0,
2573
+ "success_selector_gap": 0.0,
2574
+ "success_support_gap": 0.0,
2575
+ "support_gap": 0.17026761174201965,
2576
+ "tau": 0.10322580228482307
2577
+ },
2578
+ "747260437/train1": {
2579
+ "base_success": 1.0,
2580
+ "base_utility": 2.0,
2581
+ "coverage": 0.0,
2582
+ "execute_generated": 0.0,
2583
+ "fallback_rate": 1.0,
2584
+ "hidden_chart_oracle_success": 0.0,
2585
+ "hidden_chart_oracle_utility": 0.2786632478237152,
2586
+ "outcome_ptr": 0.0,
2587
+ "predicted_margin": -0.042144801822074585,
2588
+ "proposal_oracle_success": 0.0,
2589
+ "proposal_oracle_utility": 0.05267423391342163,
2590
+ "selected_candidate_index": 3.0,
2591
+ "selected_success": 1.0,
2592
+ "selected_success_gain_over_base": 0.0,
2593
+ "selected_utility": 2.0,
2594
+ "selected_utility_gain_over_base": 0.0,
2595
+ "selector_regret": 0.0,
2596
+ "success_selector_gap": 0.0,
2597
+ "success_support_gap": 0.0,
2598
+ "support_gap": 0.22598901391029358,
2599
+ "tau": 0.10322580228482307
2600
+ },
2601
+ "747260437/train2": {
2602
+ "base_success": 1.0,
2603
+ "base_utility": 2.0,
2604
+ "coverage": 0.0,
2605
+ "execute_generated": 0.0,
2606
+ "fallback_rate": 1.0,
2607
+ "hidden_chart_oracle_success": 0.0,
2608
+ "hidden_chart_oracle_utility": 0.2786632478237152,
2609
+ "outcome_ptr": 0.0,
2610
+ "predicted_margin": -0.04082264738847724,
2611
+ "proposal_oracle_success": 0.0,
2612
+ "proposal_oracle_utility": 0.12510187923908234,
2613
+ "selected_candidate_index": 4.0,
2614
+ "selected_success": 1.0,
2615
+ "selected_success_gain_over_base": 0.0,
2616
+ "selected_utility": 2.0,
2617
+ "selected_utility_gain_over_base": 0.0,
2618
+ "selector_regret": 0.0,
2619
+ "success_selector_gap": 0.0,
2620
+ "success_support_gap": 0.0,
2621
+ "support_gap": 0.15356136858463287,
2622
+ "tau": 0.10322580228482307
2623
+ },
2624
+ "750448296/train0": {
2625
+ "base_success": 0.0,
2626
+ "base_utility": 0.11759314934412639,
2627
+ "coverage": 0.3333333333333333,
2628
+ "execute_generated": 0.3333333333333333,
2629
+ "fallback_rate": 0.6666666666666666,
2630
+ "hidden_chart_oracle_success": 0.6666666666666666,
2631
+ "hidden_chart_oracle_utility": 1.4079964955647786,
2632
+ "outcome_ptr": 0.6666666666666666,
2633
+ "predicted_margin": -0.13841695448501412,
2634
+ "proposal_oracle_success": 0.3333333333333333,
2635
+ "proposal_oracle_utility": 0.7526603279014429,
2636
+ "selected_candidate_index": 2.3333333333333335,
2637
+ "selected_success": 0.0,
2638
+ "selected_success_gain_over_base": 0.0,
2639
+ "selected_utility": 0.1003921168545882,
2640
+ "selected_utility_gain_over_base": -0.017201032489538193,
2641
+ "selector_regret": 0.6522682110468546,
2642
+ "success_selector_gap": 0.3333333333333333,
2643
+ "success_support_gap": 0.3333333333333333,
2644
+ "support_gap": 0.6553361676633358,
2645
+ "tau": 0.10322580228482307
2646
+ },
2647
+ "750448296/train1": {
2648
+ "base_success": 0.0,
2649
+ "base_utility": 0.11759314934412639,
2650
+ "coverage": 0.0,
2651
+ "execute_generated": 0.0,
2652
+ "fallback_rate": 1.0,
2653
+ "hidden_chart_oracle_success": 0.6666666666666666,
2654
+ "hidden_chart_oracle_utility": 1.4079964955647786,
2655
+ "outcome_ptr": 1.0,
2656
+ "predicted_margin": -0.1954640674869561,
2657
+ "proposal_oracle_success": 0.6666666666666666,
2658
+ "proposal_oracle_utility": 1.3705032741030057,
2659
+ "selected_candidate_index": 4.0,
2660
+ "selected_success": 0.0,
2661
+ "selected_success_gain_over_base": 0.0,
2662
+ "selected_utility": 0.11759314934412639,
2663
+ "selected_utility_gain_over_base": 0.0,
2664
+ "selector_regret": 1.2529101247588794,
2665
+ "success_selector_gap": 0.6666666666666666,
2666
+ "success_support_gap": 0.0,
2667
+ "support_gap": 0.03749322146177292,
2668
+ "tau": 0.10322580228482307
2669
+ },
2670
+ "750448296/train2": {
2671
+ "base_success": 0.0,
2672
+ "base_utility": 0.11759314934412639,
2673
+ "coverage": 0.0,
2674
+ "execute_generated": 0.0,
2675
+ "fallback_rate": 1.0,
2676
+ "hidden_chart_oracle_success": 0.6666666666666666,
2677
+ "hidden_chart_oracle_utility": 1.4079964955647786,
2678
+ "outcome_ptr": 0.6666666666666666,
2679
+ "predicted_margin": -0.20711323699691467,
2680
+ "proposal_oracle_success": 0.6666666666666666,
2681
+ "proposal_oracle_utility": 1.349846323331197,
2682
+ "selected_candidate_index": 2.3333333333333335,
2683
+ "selected_success": 0.0,
2684
+ "selected_success_gain_over_base": 0.0,
2685
+ "selected_utility": 0.11759314934412639,
2686
+ "selected_utility_gain_over_base": 0.0,
2687
+ "selector_regret": 1.246739347775777,
2688
+ "success_selector_gap": 0.6666666666666666,
2689
+ "success_support_gap": 0.0,
2690
+ "support_gap": 0.05815017223358154,
2691
+ "tau": 0.10322580228482307
2692
+ },
2693
+ "767747424/train0": {
2694
+ "base_success": 0.25,
2695
+ "base_utility": 0.5731993420049548,
2696
+ "coverage": 0.0,
2697
+ "execute_generated": 0.0,
2698
+ "fallback_rate": 1.0,
2699
+ "hidden_chart_oracle_success": 0.25,
2700
+ "hidden_chart_oracle_utility": 0.6737305596470833,
2701
+ "outcome_ptr": 0.5,
2702
+ "predicted_margin": -0.07966992945683483,
2703
+ "proposal_oracle_success": 0.5,
2704
+ "proposal_oracle_utility": 1.0424291878007352,
2705
+ "selected_candidate_index": 3.25,
2706
+ "selected_success": 0.25,
2707
+ "selected_success_gain_over_base": 0.0,
2708
+ "selected_utility": 0.5731993420049548,
2709
+ "selected_utility_gain_over_base": 0.0,
2710
+ "selector_regret": 0.4779601227492094,
2711
+ "success_selector_gap": 0.25,
2712
+ "success_support_gap": 0.0,
2713
+ "support_gap": 0.06562284985557199,
2714
+ "tau": 0.10322580228482307
2715
+ },
2716
+ "767747424/train1": {
2717
+ "base_success": 0.25,
2718
+ "base_utility": 0.5731993420049548,
2719
+ "coverage": 0.0,
2720
+ "execute_generated": 0.0,
2721
+ "fallback_rate": 1.0,
2722
+ "hidden_chart_oracle_success": 0.25,
2723
+ "hidden_chart_oracle_utility": 0.6737305596470833,
2724
+ "outcome_ptr": 0.25,
2725
+ "predicted_margin": -0.034906867833017746,
2726
+ "proposal_oracle_success": 0.25,
2727
+ "proposal_oracle_utility": 0.5951075805351138,
2728
+ "selected_candidate_index": 2.75,
2729
+ "selected_success": 0.25,
2730
+ "selected_success_gain_over_base": 0.0,
2731
+ "selected_utility": 0.5731993420049548,
2732
+ "selected_utility_gain_over_base": 0.0,
2733
+ "selector_regret": 0.03301048465073109,
2734
+ "success_selector_gap": 0.0,
2735
+ "success_support_gap": 0.25,
2736
+ "support_gap": 0.5129444571211934,
2737
+ "tau": 0.10322580228482307
2738
+ },
2739
+ "767747424/train2": {
2740
+ "base_success": 0.25,
2741
+ "base_utility": 0.5731993420049548,
2742
+ "coverage": 0.0,
2743
+ "execute_generated": 0.0,
2744
+ "fallback_rate": 1.0,
2745
+ "hidden_chart_oracle_success": 0.25,
2746
+ "hidden_chart_oracle_utility": 0.6737305596470833,
2747
+ "outcome_ptr": 0.5,
2748
+ "predicted_margin": 0.003745219265902381,
2749
+ "proposal_oracle_success": 0.25,
2750
+ "proposal_oracle_utility": 0.6134321661666036,
2751
+ "selected_candidate_index": 3.0,
2752
+ "selected_success": 0.25,
2753
+ "selected_success_gain_over_base": 0.0,
2754
+ "selected_utility": 0.5731993420049548,
2755
+ "selected_utility_gain_over_base": 0.0,
2756
+ "selector_regret": 0.04225692339241505,
2757
+ "success_selector_gap": 0.0,
2758
+ "success_support_gap": 0.25,
2759
+ "support_gap": 0.49461987148970366,
2760
+ "tau": 0.10322580228482307
2761
+ }
2762
+ }
workspace/runs/ctt_base_context_obs_learned_dominance_basic_envclip_train_to_test/metrics_by_task.json ADDED
@@ -0,0 +1,117 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "LiftPegUpright-v1": {
3
+ "base_success": 0.2,
4
+ "base_utility": 0.9498705387115478,
5
+ "coverage": 0.06666666666666667,
6
+ "execute_generated": 0.06666666666666667,
7
+ "fallback_rate": 0.9333333333333333,
8
+ "hidden_chart_oracle_success": 1.0,
9
+ "hidden_chart_oracle_utility": 2.0,
10
+ "outcome_ptr": 0.4666666666666667,
11
+ "predicted_margin": 0.05400229400248756,
12
+ "proposal_oracle_success": 0.4666666666666667,
13
+ "proposal_oracle_utility": 1.2244940986235937,
14
+ "selected_candidate_index": 3.2666666666666666,
15
+ "selected_success": 0.2,
16
+ "selected_success_gain_over_base": 0.0,
17
+ "selected_utility": 0.9233068734407425,
18
+ "selected_utility_gain_over_base": -0.026563665270805357,
19
+ "selector_regret": 0.42827867666880287,
20
+ "success_selector_gap": 0.3333333333333333,
21
+ "success_support_gap": 0.5333333333333333,
22
+ "support_gap": 0.7755059013764064,
23
+ "tau": 0.10322580228482309
24
+ },
25
+ "PickCube-v1": {
26
+ "base_success": 0.0,
27
+ "base_utility": 0.5839591026306152,
28
+ "coverage": 0.0,
29
+ "execute_generated": 0.0,
30
+ "fallback_rate": 1.0,
31
+ "hidden_chart_oracle_success": 1.0,
32
+ "hidden_chart_oracle_utility": 2.0,
33
+ "outcome_ptr": 0.0,
34
+ "predicted_margin": -0.14277515452222114,
35
+ "proposal_oracle_success": 0.0,
36
+ "proposal_oracle_utility": 0.13425676027933756,
37
+ "selected_candidate_index": 4.0,
38
+ "selected_success": 0.0,
39
+ "selected_success_gain_over_base": 0.0,
40
+ "selected_utility": 0.5839591026306152,
41
+ "selected_utility_gain_over_base": 0.0,
42
+ "selector_regret": 0.0,
43
+ "success_selector_gap": 0.0,
44
+ "success_support_gap": 1.0,
45
+ "support_gap": 1.8657432397206624,
46
+ "tau": 0.10322580228482307
47
+ },
48
+ "PullCube-v1": {
49
+ "base_success": 0.36,
50
+ "base_utility": 0.7751704612234608,
51
+ "coverage": 0.18666666666666668,
52
+ "execute_generated": 0.18666666666666668,
53
+ "fallback_rate": 0.8133333333333334,
54
+ "hidden_chart_oracle_success": 0.52,
55
+ "hidden_chart_oracle_utility": 1.1467878860235214,
56
+ "outcome_ptr": 0.44,
57
+ "predicted_margin": -0.009254413460779327,
58
+ "proposal_oracle_success": 0.4666666666666667,
59
+ "proposal_oracle_utility": 0.9816135169876119,
60
+ "selected_candidate_index": 2.9466666666666668,
61
+ "selected_success": 0.36,
62
+ "selected_success_gain_over_base": 0.0,
63
+ "selected_utility": 0.7723906052314365,
64
+ "selected_utility_gain_over_base": -0.0027798559920241437,
65
+ "selector_regret": 0.3914262486419951,
66
+ "success_selector_gap": 0.2,
67
+ "success_support_gap": 0.09333333333333334,
68
+ "support_gap": 0.23527305972451965,
69
+ "tau": 0.1032258022848229
70
+ },
71
+ "PushCube-v1": {
72
+ "base_success": 0.16666666666666666,
73
+ "base_utility": 0.3924041787783305,
74
+ "coverage": 0.6666666666666666,
75
+ "execute_generated": 0.6666666666666666,
76
+ "fallback_rate": 0.3333333333333333,
77
+ "hidden_chart_oracle_success": 0.8333333333333334,
78
+ "hidden_chart_oracle_utility": 1.7015800600250561,
79
+ "outcome_ptr": 0.7777777777777778,
80
+ "predicted_margin": 0.1337068728029153,
81
+ "proposal_oracle_success": 0.6666666666666666,
82
+ "proposal_oracle_utility": 1.365773618221283,
83
+ "selected_candidate_index": 2.8333333333333335,
84
+ "selected_success": 0.3333333333333333,
85
+ "selected_success_gain_over_base": 0.16666666666666666,
86
+ "selected_utility": 0.7242705598473549,
87
+ "selected_utility_gain_over_base": 0.3318663810690244,
88
+ "selector_regret": 0.642994626528687,
89
+ "success_selector_gap": 0.3333333333333333,
90
+ "success_support_gap": 0.3333333333333333,
91
+ "support_gap": 0.634226381778717,
92
+ "tau": 0.10322580228482309
93
+ },
94
+ "StackCube-v1": {
95
+ "base_success": 0.2727272727272727,
96
+ "base_utility": 0.7763947045261209,
97
+ "coverage": 0.15151515151515152,
98
+ "execute_generated": 0.15151515151515152,
99
+ "fallback_rate": 0.8484848484848485,
100
+ "hidden_chart_oracle_success": 1.0,
101
+ "hidden_chart_oracle_utility": 2.0,
102
+ "outcome_ptr": 0.5757575757575758,
103
+ "predicted_margin": 0.042076341440502645,
104
+ "proposal_oracle_success": 0.45454545454545453,
105
+ "proposal_oracle_utility": 1.0546941955884297,
106
+ "selected_candidate_index": 3.4545454545454546,
107
+ "selected_success": 0.30303030303030304,
108
+ "selected_success_gain_over_base": 0.030303030303030304,
109
+ "selected_utility": 0.8081388062599933,
110
+ "selected_utility_gain_over_base": 0.03174410173387238,
111
+ "selector_regret": 0.25066923237208166,
112
+ "success_selector_gap": 0.15151515151515152,
113
+ "success_support_gap": 0.5454545454545454,
114
+ "support_gap": 0.9453058044115702,
115
+ "tau": 0.10322580228482311
116
+ }
117
+ }
workspace/runs/ctt_base_context_obs_learned_dominance_basic_envclip_train_to_test/report.md ADDED
@@ -0,0 +1,17 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ # Learned Dominance-Calibrated CTT Selector
2
+
3
+ Calibration rows: `432`
4
+ Eval rows: `144`
5
+ Selected ridge lambda: `100.0`
6
+ Tau: `0.103226`
7
+ Feature set: `basic`
8
+ Target: `utility_margin`
9
+
10
+ The ridge calibrator and threshold are fit on calibration measured rows only. Eval outcomes are used only for reporting.
11
+
12
+ | Split | Coverage | Fallback | Base success | Selected success | Proposal oracle | OutcomePTR | Success support gap | Success selector gap |
13
+ | --- | ---: | ---: | ---: | ---: | ---: | ---: | ---: | ---: |
14
+ | calibration | 0.3241 | 0.6759 | 0.2778 | 0.3241 | 0.4722 | 0.4907 | 0.2315 | 0.2083 |
15
+ | eval | 0.2222 | 0.7778 | 0.2917 | 0.3194 | 0.4792 | 0.5069 | 0.2917 | 0.2153 |
16
+
17
+ This is a selector diagnostic over already measured candidates, not a new rollout.
workspace/runs/ctt_base_context_obs_learned_dominance_basic_envclip_train_to_test/split_hash.txt ADDED
@@ -0,0 +1,14 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "calibration_target_index": {
3
+ "content_hash": "67f7b4a692f7c7e71da24378dc71e3399c5d35afc14c494869eba6087b765c42",
4
+ "retrieval_index_allowed": true,
5
+ "split": "train",
6
+ "split_hash": "867a94a1c67686c082765671a7bf8d74a9d633085e1b2619c83197c60bf58372"
7
+ },
8
+ "eval_target_index": {
9
+ "content_hash": "0c38d73dc9bc4ab756e12006671b5d12f1a23358b281e54b3f6a9ae3cbf6e9af",
10
+ "retrieval_index_allowed": false,
11
+ "split": "test",
12
+ "split_hash": "84151df9e88c219b76751677558cc5bbbc19082bf358cea7330158e9e403e25f"
13
+ }
14
+ }
workspace/runs/ctt_base_context_obs_learned_dominance_basic_envclip_train_to_test/table.tex ADDED
@@ -0,0 +1,8 @@
 
 
 
 
 
 
 
 
 
1
+ % Auto-generated by scripts/eval_learned_dominance_selector.py
2
+ \begin{tabular}{lrrrrrrrr}
3
+ \toprule
4
+ Rows & Coverage & Fallback & Base succ. & Selected succ. & Oracle succ. & OutcomePTR & Succ. support gap & Succ. selector gap \\
5
+ \midrule
6
+ 144 & 0.2222 & 0.7778 & 0.2917 & 0.3194 & 0.4792 & 0.5069 & 0.2917 & 0.2153 \\
7
+ \bottomrule
8
+ \end{tabular}
workspace/runs/ctt_base_context_obs_learned_dominance_basic_envclip_train_to_test/train.log ADDED
@@ -0,0 +1,4 @@
 
 
 
 
 
1
+ trained ridge dominance calibrator on calibration measured rows only
2
+ selected_lambda=100.0
3
+ tau=0.103226
4
+ calibration_input=runs/ctt_base_context_obs_train_cal_envclip_rollout_comparison/combined_measured_candidates.json
workspace/runs/ctt_base_context_obs_learned_dominance_context_envclip_train_to_test/command.txt ADDED
@@ -0,0 +1 @@
 
 
1
+ python scripts/eval_learned_dominance_selector.py --calibration-input runs/ctt_base_context_obs_train_cal_envclip_rollout_comparison/combined_measured_candidates.json --calibration-target-index data/cil_charts_rgb_refs/train/index.json --eval-input runs/ctt_base_context_obs_test_envclip_rollout_comparison/combined_measured_candidates.json --eval-target-index data/cil_charts_rgb_refs/test/index.json --checkpoint-template runs/ctt_residual_base_context_obs_seed{seed}/model.pt --out-dir runs/ctt_base_context_obs_learned_dominance_context_envclip_train_to_test --feature-set context --target utility_margin --k 8 --bootstrap-samples 1000
workspace/runs/ctt_base_context_obs_learned_dominance_context_envclip_train_to_test/config.yaml ADDED
@@ -0,0 +1,11 @@
 
 
 
 
 
 
 
 
 
 
 
 
1
+ bootstrap_samples: 1000
2
+ calibration_input: runs/ctt_base_context_obs_train_cal_envclip_rollout_comparison/combined_measured_candidates.json
3
+ calibration_target_index: data/cil_charts_rgb_refs/train/index.json
4
+ checkpoint_template: runs/ctt_residual_base_context_obs_seed{seed}/model.pt
5
+ eval_input: runs/ctt_base_context_obs_test_envclip_rollout_comparison/combined_measured_candidates.json
6
+ eval_target_index: data/cil_charts_rgb_refs/test/index.json
7
+ feature_set: context
8
+ k: 8
9
+ out_dir: runs/ctt_base_context_obs_learned_dominance_context_envclip_train_to_test
10
+ ridge_lambdas: 0,0.01,0.1,1,10,100
11
+ target: utility_margin
workspace/runs/ctt_base_context_obs_learned_dominance_context_envclip_train_to_test/data_hash.txt ADDED
@@ -0,0 +1,6 @@
 
 
 
 
 
 
 
1
+ {
2
+ "calibration_input": "06375099001e740a377c402e34b77bf6974be3dc2a468974340ce41eb1b487f0",
3
+ "calibration_target_index": "67973c50072aa87a3b97533b6292b08a8b4d88a95571dc22e91b6e8fd19e0d85",
4
+ "eval_input": "57df43ff4802379dfd60b5209d5a04028d721256406685decededccdbec0d49c",
5
+ "eval_target_index": "dee1b5225485df6624747dd4c5074bcf514685b54017d05e698d2051638fbf36"
6
+ }
workspace/runs/ctt_base_context_obs_learned_dominance_context_envclip_train_to_test/eval.log ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ evaluated learned fallback rule on held-out measured rollout rows
2
+ eval_input=runs/ctt_base_context_obs_test_envclip_rollout_comparison/combined_measured_candidates.json
3
+ num_eval_rows=144
workspace/runs/ctt_base_context_obs_learned_dominance_context_envclip_train_to_test/git_hash.txt ADDED
@@ -0,0 +1 @@
 
 
1
+ 14e81292e2004925f81e0bb4810f6f144bfd87b7
workspace/runs/ctt_base_context_obs_learned_dominance_context_envclip_train_to_test/metrics.json ADDED
The diff for this file is too large to render. See raw diff
 
workspace/runs/ctt_base_context_obs_learned_dominance_context_envclip_train_to_test/metrics_by_seed.json ADDED
@@ -0,0 +1,2762 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "1073161778/train0": {
3
+ "base_success": 1.0,
4
+ "base_utility": 2.0,
5
+ "coverage": 0.0,
6
+ "execute_generated": 0.0,
7
+ "fallback_rate": 1.0,
8
+ "hidden_chart_oracle_success": 1.0,
9
+ "hidden_chart_oracle_utility": 2.0,
10
+ "outcome_ptr": 0.0,
11
+ "predicted_margin": 0.03594391742610452,
12
+ "proposal_oracle_success": 1.0,
13
+ "proposal_oracle_utility": 2.0,
14
+ "selected_candidate_index": 4.0,
15
+ "selected_success": 1.0,
16
+ "selected_success_gain_over_base": 0.0,
17
+ "selected_utility": 2.0,
18
+ "selected_utility_gain_over_base": 0.0,
19
+ "selector_regret": 0.0,
20
+ "success_selector_gap": 0.0,
21
+ "success_support_gap": 0.0,
22
+ "support_gap": 0.0,
23
+ "tau": 0.08716187110665122
24
+ },
25
+ "1073161778/train1": {
26
+ "base_success": 1.0,
27
+ "base_utility": 2.0,
28
+ "coverage": 0.0,
29
+ "execute_generated": 0.0,
30
+ "fallback_rate": 1.0,
31
+ "hidden_chart_oracle_success": 1.0,
32
+ "hidden_chart_oracle_utility": 2.0,
33
+ "outcome_ptr": 0.0,
34
+ "predicted_margin": -0.04561956712124951,
35
+ "proposal_oracle_success": 0.0,
36
+ "proposal_oracle_utility": 0.7071996331214905,
37
+ "selected_candidate_index": 3.0,
38
+ "selected_success": 1.0,
39
+ "selected_success_gain_over_base": 0.0,
40
+ "selected_utility": 2.0,
41
+ "selected_utility_gain_over_base": 0.0,
42
+ "selector_regret": 0.0,
43
+ "success_selector_gap": 0.0,
44
+ "success_support_gap": 1.0,
45
+ "support_gap": 1.2928003668785095,
46
+ "tau": 0.08716187110665122
47
+ },
48
+ "1073161778/train2": {
49
+ "base_success": 1.0,
50
+ "base_utility": 2.0,
51
+ "coverage": 0.0,
52
+ "execute_generated": 0.0,
53
+ "fallback_rate": 1.0,
54
+ "hidden_chart_oracle_success": 1.0,
55
+ "hidden_chart_oracle_utility": 2.0,
56
+ "outcome_ptr": 0.0,
57
+ "predicted_margin": 0.028138900652246862,
58
+ "proposal_oracle_success": 1.0,
59
+ "proposal_oracle_utility": 2.0,
60
+ "selected_candidate_index": 2.0,
61
+ "selected_success": 1.0,
62
+ "selected_success_gain_over_base": 0.0,
63
+ "selected_utility": 2.0,
64
+ "selected_utility_gain_over_base": 0.0,
65
+ "selector_regret": 0.0,
66
+ "success_selector_gap": 0.0,
67
+ "success_support_gap": 0.0,
68
+ "support_gap": 0.0,
69
+ "tau": 0.08716187110665122
70
+ },
71
+ "1113371261/train0": {
72
+ "base_success": 0.0,
73
+ "base_utility": 0.15842342376708984,
74
+ "coverage": 0.0,
75
+ "execute_generated": 0.0,
76
+ "fallback_rate": 1.0,
77
+ "hidden_chart_oracle_success": 1.0,
78
+ "hidden_chart_oracle_utility": 2.0,
79
+ "outcome_ptr": 0.0,
80
+ "predicted_margin": -0.00836323376300073,
81
+ "proposal_oracle_success": 0.0,
82
+ "proposal_oracle_utility": 0.10274861752986908,
83
+ "selected_candidate_index": 1.0,
84
+ "selected_success": 0.0,
85
+ "selected_success_gain_over_base": 0.0,
86
+ "selected_utility": 0.15842342376708984,
87
+ "selected_utility_gain_over_base": 0.0,
88
+ "selector_regret": 0.0,
89
+ "success_selector_gap": 0.0,
90
+ "success_support_gap": 1.0,
91
+ "support_gap": 1.897251382470131,
92
+ "tau": 0.08716187110665122
93
+ },
94
+ "1113371261/train1": {
95
+ "base_success": 0.0,
96
+ "base_utility": 0.15842342376708984,
97
+ "coverage": 0.0,
98
+ "execute_generated": 0.0,
99
+ "fallback_rate": 1.0,
100
+ "hidden_chart_oracle_success": 1.0,
101
+ "hidden_chart_oracle_utility": 2.0,
102
+ "outcome_ptr": 1.0,
103
+ "predicted_margin": 0.044851987296773074,
104
+ "proposal_oracle_success": 0.0,
105
+ "proposal_oracle_utility": 0.18093222379684448,
106
+ "selected_candidate_index": 2.0,
107
+ "selected_success": 0.0,
108
+ "selected_success_gain_over_base": 0.0,
109
+ "selected_utility": 0.15842342376708984,
110
+ "selected_utility_gain_over_base": 0.0,
111
+ "selector_regret": 0.02250880002975464,
112
+ "success_selector_gap": 0.0,
113
+ "success_support_gap": 1.0,
114
+ "support_gap": 1.8190677762031555,
115
+ "tau": 0.08716187110665122
116
+ },
117
+ "1113371261/train2": {
118
+ "base_success": 0.0,
119
+ "base_utility": 0.15842342376708984,
120
+ "coverage": 0.0,
121
+ "execute_generated": 0.0,
122
+ "fallback_rate": 1.0,
123
+ "hidden_chart_oracle_success": 1.0,
124
+ "hidden_chart_oracle_utility": 2.0,
125
+ "outcome_ptr": 1.0,
126
+ "predicted_margin": -0.01822403326470206,
127
+ "proposal_oracle_success": 0.0,
128
+ "proposal_oracle_utility": 0.19527167081832886,
129
+ "selected_candidate_index": 4.0,
130
+ "selected_success": 0.0,
131
+ "selected_success_gain_over_base": 0.0,
132
+ "selected_utility": 0.15842342376708984,
133
+ "selected_utility_gain_over_base": 0.0,
134
+ "selector_regret": 0.036848247051239014,
135
+ "success_selector_gap": 0.0,
136
+ "success_support_gap": 1.0,
137
+ "support_gap": 1.8047283291816711,
138
+ "tau": 0.08716187110665122
139
+ },
140
+ "1151410346/train0": {
141
+ "base_success": 0.0,
142
+ "base_utility": 0.2260012924671173,
143
+ "coverage": 1.0,
144
+ "execute_generated": 1.0,
145
+ "fallback_rate": 0.0,
146
+ "hidden_chart_oracle_success": 1.0,
147
+ "hidden_chart_oracle_utility": 2.0,
148
+ "outcome_ptr": 1.0,
149
+ "predicted_margin": 0.11912427265923112,
150
+ "proposal_oracle_success": 0.0,
151
+ "proposal_oracle_utility": 0.8663521409034729,
152
+ "selected_candidate_index": 5.0,
153
+ "selected_success": 0.0,
154
+ "selected_success_gain_over_base": 0.0,
155
+ "selected_utility": 0.18858957290649414,
156
+ "selected_utility_gain_over_base": -0.03741171956062317,
157
+ "selector_regret": 0.6777625679969788,
158
+ "success_selector_gap": 0.0,
159
+ "success_support_gap": 1.0,
160
+ "support_gap": 1.133647859096527,
161
+ "tau": 0.08716187110665122
162
+ },
163
+ "1151410346/train1": {
164
+ "base_success": 0.0,
165
+ "base_utility": 0.2260012924671173,
166
+ "coverage": 1.0,
167
+ "execute_generated": 1.0,
168
+ "fallback_rate": 0.0,
169
+ "hidden_chart_oracle_success": 1.0,
170
+ "hidden_chart_oracle_utility": 2.0,
171
+ "outcome_ptr": 1.0,
172
+ "predicted_margin": 0.12531527657951597,
173
+ "proposal_oracle_success": 0.0,
174
+ "proposal_oracle_utility": 0.8138919472694397,
175
+ "selected_candidate_index": 1.0,
176
+ "selected_success": 0.0,
177
+ "selected_success_gain_over_base": 0.0,
178
+ "selected_utility": 0.812879204750061,
179
+ "selected_utility_gain_over_base": 0.5868779122829437,
180
+ "selector_regret": 0.0010127425193786621,
181
+ "success_selector_gap": 0.0,
182
+ "success_support_gap": 1.0,
183
+ "support_gap": 1.1861080527305603,
184
+ "tau": 0.08716187110665122
185
+ },
186
+ "1151410346/train2": {
187
+ "base_success": 0.0,
188
+ "base_utility": 0.2260012924671173,
189
+ "coverage": 1.0,
190
+ "execute_generated": 1.0,
191
+ "fallback_rate": 0.0,
192
+ "hidden_chart_oracle_success": 1.0,
193
+ "hidden_chart_oracle_utility": 2.0,
194
+ "outcome_ptr": 0.0,
195
+ "predicted_margin": 0.14052310673958157,
196
+ "proposal_oracle_success": 0.0,
197
+ "proposal_oracle_utility": 0.21912741661071777,
198
+ "selected_candidate_index": 4.0,
199
+ "selected_success": 0.0,
200
+ "selected_success_gain_over_base": 0.0,
201
+ "selected_utility": 0.2044079303741455,
202
+ "selected_utility_gain_over_base": -0.021593362092971802,
203
+ "selector_regret": 0.014719486236572266,
204
+ "success_selector_gap": 0.0,
205
+ "success_support_gap": 1.0,
206
+ "support_gap": 1.7808725833892822,
207
+ "tau": 0.08716187110665122
208
+ },
209
+ "1261930456/train0": {
210
+ "base_success": 1.0,
211
+ "base_utility": 2.0,
212
+ "coverage": 1.0,
213
+ "execute_generated": 1.0,
214
+ "fallback_rate": 0.0,
215
+ "hidden_chart_oracle_success": 1.0,
216
+ "hidden_chart_oracle_utility": 2.0,
217
+ "outcome_ptr": 0.0,
218
+ "predicted_margin": 0.19659137771964635,
219
+ "proposal_oracle_success": 1.0,
220
+ "proposal_oracle_utility": 2.0,
221
+ "selected_candidate_index": 3.0,
222
+ "selected_success": 0.0,
223
+ "selected_success_gain_over_base": -1.0,
224
+ "selected_utility": 0.06716175377368927,
225
+ "selected_utility_gain_over_base": -1.9328382462263107,
226
+ "selector_regret": 1.9328382462263107,
227
+ "success_selector_gap": 1.0,
228
+ "success_support_gap": 0.0,
229
+ "support_gap": 0.0,
230
+ "tau": 0.08716187110665122
231
+ },
232
+ "1261930456/train1": {
233
+ "base_success": 1.0,
234
+ "base_utility": 2.0,
235
+ "coverage": 1.0,
236
+ "execute_generated": 1.0,
237
+ "fallback_rate": 0.0,
238
+ "hidden_chart_oracle_success": 1.0,
239
+ "hidden_chart_oracle_utility": 2.0,
240
+ "outcome_ptr": 0.0,
241
+ "predicted_margin": 0.24580152506458247,
242
+ "proposal_oracle_success": 1.0,
243
+ "proposal_oracle_utility": 2.0,
244
+ "selected_candidate_index": 3.0,
245
+ "selected_success": 0.0,
246
+ "selected_success_gain_over_base": -1.0,
247
+ "selected_utility": 0.0566796213388443,
248
+ "selected_utility_gain_over_base": -1.9433203786611557,
249
+ "selector_regret": 1.9433203786611557,
250
+ "success_selector_gap": 1.0,
251
+ "success_support_gap": 0.0,
252
+ "support_gap": 0.0,
253
+ "tau": 0.08716187110665122
254
+ },
255
+ "1261930456/train2": {
256
+ "base_success": 1.0,
257
+ "base_utility": 2.0,
258
+ "coverage": 1.0,
259
+ "execute_generated": 1.0,
260
+ "fallback_rate": 0.0,
261
+ "hidden_chart_oracle_success": 1.0,
262
+ "hidden_chart_oracle_utility": 2.0,
263
+ "outcome_ptr": 0.0,
264
+ "predicted_margin": 0.19688646509179342,
265
+ "proposal_oracle_success": 1.0,
266
+ "proposal_oracle_utility": 2.0,
267
+ "selected_candidate_index": 4.0,
268
+ "selected_success": 0.0,
269
+ "selected_success_gain_over_base": -1.0,
270
+ "selected_utility": 0.08501242101192474,
271
+ "selected_utility_gain_over_base": -1.9149875789880753,
272
+ "selector_regret": 1.9149875789880753,
273
+ "success_selector_gap": 1.0,
274
+ "success_support_gap": 0.0,
275
+ "support_gap": 0.0,
276
+ "tau": 0.08716187110665122
277
+ },
278
+ "1324133122/train0": {
279
+ "base_success": 0.0,
280
+ "base_utility": 0.09220491349697113,
281
+ "coverage": 0.0,
282
+ "execute_generated": 0.0,
283
+ "fallback_rate": 1.0,
284
+ "hidden_chart_oracle_success": 1.0,
285
+ "hidden_chart_oracle_utility": 2.0,
286
+ "outcome_ptr": 0.0,
287
+ "predicted_margin": -0.06560063067574379,
288
+ "proposal_oracle_success": 0.0,
289
+ "proposal_oracle_utility": 0.054989203810691833,
290
+ "selected_candidate_index": 2.0,
291
+ "selected_success": 0.0,
292
+ "selected_success_gain_over_base": 0.0,
293
+ "selected_utility": 0.09220491349697113,
294
+ "selected_utility_gain_over_base": 0.0,
295
+ "selector_regret": 0.0,
296
+ "success_selector_gap": 0.0,
297
+ "success_support_gap": 1.0,
298
+ "support_gap": 1.9450107961893082,
299
+ "tau": 0.08716187110665122
300
+ },
301
+ "1324133122/train1": {
302
+ "base_success": 0.0,
303
+ "base_utility": 0.09220491349697113,
304
+ "coverage": 0.0,
305
+ "execute_generated": 0.0,
306
+ "fallback_rate": 1.0,
307
+ "hidden_chart_oracle_success": 1.0,
308
+ "hidden_chart_oracle_utility": 2.0,
309
+ "outcome_ptr": 1.0,
310
+ "predicted_margin": 0.08092824613993685,
311
+ "proposal_oracle_success": 0.0,
312
+ "proposal_oracle_utility": 0.1324586570262909,
313
+ "selected_candidate_index": 6.0,
314
+ "selected_success": 0.0,
315
+ "selected_success_gain_over_base": 0.0,
316
+ "selected_utility": 0.09220491349697113,
317
+ "selected_utility_gain_over_base": 0.0,
318
+ "selector_regret": 0.04025374352931976,
319
+ "success_selector_gap": 0.0,
320
+ "success_support_gap": 1.0,
321
+ "support_gap": 1.867541342973709,
322
+ "tau": 0.08716187110665122
323
+ },
324
+ "1324133122/train2": {
325
+ "base_success": 0.0,
326
+ "base_utility": 0.09220491349697113,
327
+ "coverage": 0.0,
328
+ "execute_generated": 0.0,
329
+ "fallback_rate": 1.0,
330
+ "hidden_chart_oracle_success": 1.0,
331
+ "hidden_chart_oracle_utility": 2.0,
332
+ "outcome_ptr": 1.0,
333
+ "predicted_margin": -0.01867128006022438,
334
+ "proposal_oracle_success": 0.0,
335
+ "proposal_oracle_utility": 0.19291165471076965,
336
+ "selected_candidate_index": 2.0,
337
+ "selected_success": 0.0,
338
+ "selected_success_gain_over_base": 0.0,
339
+ "selected_utility": 0.09220491349697113,
340
+ "selected_utility_gain_over_base": 0.0,
341
+ "selector_regret": 0.10070674121379852,
342
+ "success_selector_gap": 0.0,
343
+ "success_support_gap": 1.0,
344
+ "support_gap": 1.8070883452892303,
345
+ "tau": 0.08716187110665122
346
+ },
347
+ "1324642411/train0": {
348
+ "base_success": 0.0,
349
+ "base_utility": 0.10647052526473999,
350
+ "coverage": 1.0,
351
+ "execute_generated": 1.0,
352
+ "fallback_rate": 0.0,
353
+ "hidden_chart_oracle_success": 1.0,
354
+ "hidden_chart_oracle_utility": 2.0,
355
+ "outcome_ptr": 1.0,
356
+ "predicted_margin": 0.2824797487730189,
357
+ "proposal_oracle_success": 0.0,
358
+ "proposal_oracle_utility": 0.17193320393562317,
359
+ "selected_candidate_index": 3.0,
360
+ "selected_success": 0.0,
361
+ "selected_success_gain_over_base": 0.0,
362
+ "selected_utility": 0.08104583621025085,
363
+ "selected_utility_gain_over_base": -0.025424689054489136,
364
+ "selector_regret": 0.09088736772537231,
365
+ "success_selector_gap": 0.0,
366
+ "success_support_gap": 1.0,
367
+ "support_gap": 1.8280667960643768,
368
+ "tau": 0.08716187110665122
369
+ },
370
+ "1324642411/train1": {
371
+ "base_success": 0.0,
372
+ "base_utility": 0.10647052526473999,
373
+ "coverage": 1.0,
374
+ "execute_generated": 1.0,
375
+ "fallback_rate": 0.0,
376
+ "hidden_chart_oracle_success": 1.0,
377
+ "hidden_chart_oracle_utility": 2.0,
378
+ "outcome_ptr": 1.0,
379
+ "predicted_margin": 0.22668750061917628,
380
+ "proposal_oracle_success": 0.0,
381
+ "proposal_oracle_utility": 0.1836909055709839,
382
+ "selected_candidate_index": 1.0,
383
+ "selected_success": 0.0,
384
+ "selected_success_gain_over_base": 0.0,
385
+ "selected_utility": 0.10393516719341278,
386
+ "selected_utility_gain_over_base": -0.0025353580713272095,
387
+ "selector_regret": 0.0797557383775711,
388
+ "success_selector_gap": 0.0,
389
+ "success_support_gap": 1.0,
390
+ "support_gap": 1.8163090944290161,
391
+ "tau": 0.08716187110665122
392
+ },
393
+ "1324642411/train2": {
394
+ "base_success": 0.0,
395
+ "base_utility": 0.10647052526473999,
396
+ "coverage": 0.0,
397
+ "execute_generated": 0.0,
398
+ "fallback_rate": 1.0,
399
+ "hidden_chart_oracle_success": 1.0,
400
+ "hidden_chart_oracle_utility": 2.0,
401
+ "outcome_ptr": 1.0,
402
+ "predicted_margin": -0.09456225854208301,
403
+ "proposal_oracle_success": 1.0,
404
+ "proposal_oracle_utility": 2.0,
405
+ "selected_candidate_index": 2.0,
406
+ "selected_success": 0.0,
407
+ "selected_success_gain_over_base": 0.0,
408
+ "selected_utility": 0.10647052526473999,
409
+ "selected_utility_gain_over_base": 0.0,
410
+ "selector_regret": 1.89352947473526,
411
+ "success_selector_gap": 1.0,
412
+ "success_support_gap": 0.0,
413
+ "support_gap": 0.0,
414
+ "tau": 0.08716187110665122
415
+ },
416
+ "1375492786/train0": {
417
+ "base_success": 1.0,
418
+ "base_utility": 2.0,
419
+ "coverage": 0.0,
420
+ "execute_generated": 0.0,
421
+ "fallback_rate": 1.0,
422
+ "hidden_chart_oracle_success": 1.0,
423
+ "hidden_chart_oracle_utility": 2.0,
424
+ "outcome_ptr": 0.0,
425
+ "predicted_margin": -0.21323227476554057,
426
+ "proposal_oracle_success": 1.0,
427
+ "proposal_oracle_utility": 2.0,
428
+ "selected_candidate_index": 7.0,
429
+ "selected_success": 1.0,
430
+ "selected_success_gain_over_base": 0.0,
431
+ "selected_utility": 2.0,
432
+ "selected_utility_gain_over_base": 0.0,
433
+ "selector_regret": 0.0,
434
+ "success_selector_gap": 0.0,
435
+ "success_support_gap": 0.0,
436
+ "support_gap": 0.0,
437
+ "tau": 0.08716187110665122
438
+ },
439
+ "1375492786/train1": {
440
+ "base_success": 1.0,
441
+ "base_utility": 2.0,
442
+ "coverage": 0.0,
443
+ "execute_generated": 0.0,
444
+ "fallback_rate": 1.0,
445
+ "hidden_chart_oracle_success": 1.0,
446
+ "hidden_chart_oracle_utility": 2.0,
447
+ "outcome_ptr": 0.0,
448
+ "predicted_margin": -0.38193079914559935,
449
+ "proposal_oracle_success": 0.0,
450
+ "proposal_oracle_utility": 0.08251869678497314,
451
+ "selected_candidate_index": 2.0,
452
+ "selected_success": 1.0,
453
+ "selected_success_gain_over_base": 0.0,
454
+ "selected_utility": 2.0,
455
+ "selected_utility_gain_over_base": 0.0,
456
+ "selector_regret": 0.0,
457
+ "success_selector_gap": 0.0,
458
+ "success_support_gap": 1.0,
459
+ "support_gap": 1.9174813032150269,
460
+ "tau": 0.08716187110665122
461
+ },
462
+ "1375492786/train2": {
463
+ "base_success": 1.0,
464
+ "base_utility": 2.0,
465
+ "coverage": 0.0,
466
+ "execute_generated": 0.0,
467
+ "fallback_rate": 1.0,
468
+ "hidden_chart_oracle_success": 1.0,
469
+ "hidden_chart_oracle_utility": 2.0,
470
+ "outcome_ptr": 0.0,
471
+ "predicted_margin": -0.2241678834898516,
472
+ "proposal_oracle_success": 1.0,
473
+ "proposal_oracle_utility": 2.0,
474
+ "selected_candidate_index": 7.0,
475
+ "selected_success": 1.0,
476
+ "selected_success_gain_over_base": 0.0,
477
+ "selected_utility": 2.0,
478
+ "selected_utility_gain_over_base": 0.0,
479
+ "selector_regret": 0.0,
480
+ "success_selector_gap": 0.0,
481
+ "success_support_gap": 0.0,
482
+ "support_gap": 0.0,
483
+ "tau": 0.08716187110665122
484
+ },
485
+ "141984615/train0": {
486
+ "base_success": 0.0,
487
+ "base_utility": 0.10283424705266953,
488
+ "coverage": 0.0,
489
+ "execute_generated": 0.0,
490
+ "fallback_rate": 1.0,
491
+ "hidden_chart_oracle_success": 0.0,
492
+ "hidden_chart_oracle_utility": 0.2857772707939148,
493
+ "outcome_ptr": 0.0,
494
+ "predicted_margin": -0.0008772076126836725,
495
+ "proposal_oracle_success": 0.0,
496
+ "proposal_oracle_utility": 0.007291515823453665,
497
+ "selected_candidate_index": 3.0,
498
+ "selected_success": 0.0,
499
+ "selected_success_gain_over_base": 0.0,
500
+ "selected_utility": 0.10283424705266953,
501
+ "selected_utility_gain_over_base": 0.0,
502
+ "selector_regret": 0.0,
503
+ "success_selector_gap": 0.0,
504
+ "success_support_gap": 0.0,
505
+ "support_gap": 0.27848575497046113,
506
+ "tau": 0.08716187110665122
507
+ },
508
+ "141984615/train1": {
509
+ "base_success": 0.0,
510
+ "base_utility": 0.10283424705266953,
511
+ "coverage": 1.0,
512
+ "execute_generated": 1.0,
513
+ "fallback_rate": 0.0,
514
+ "hidden_chart_oracle_success": 0.0,
515
+ "hidden_chart_oracle_utility": 0.2857772707939148,
516
+ "outcome_ptr": 0.0,
517
+ "predicted_margin": 0.09693487495218861,
518
+ "proposal_oracle_success": 0.0,
519
+ "proposal_oracle_utility": 0.037011127918958664,
520
+ "selected_candidate_index": 1.0,
521
+ "selected_success": 0.0,
522
+ "selected_success_gain_over_base": 0.0,
523
+ "selected_utility": 0.027254482731223106,
524
+ "selected_utility_gain_over_base": -0.07557976432144642,
525
+ "selector_regret": 0.009756645187735558,
526
+ "success_selector_gap": 0.0,
527
+ "success_support_gap": 0.0,
528
+ "support_gap": 0.24876614287495613,
529
+ "tau": 0.08716187110665122
530
+ },
531
+ "141984615/train2": {
532
+ "base_success": 0.0,
533
+ "base_utility": 0.10283424705266953,
534
+ "coverage": 0.0,
535
+ "execute_generated": 0.0,
536
+ "fallback_rate": 1.0,
537
+ "hidden_chart_oracle_success": 0.0,
538
+ "hidden_chart_oracle_utility": 0.2857772707939148,
539
+ "outcome_ptr": 0.0,
540
+ "predicted_margin": -0.13186726337212987,
541
+ "proposal_oracle_success": 0.0,
542
+ "proposal_oracle_utility": 0.030228236690163612,
543
+ "selected_candidate_index": 4.0,
544
+ "selected_success": 0.0,
545
+ "selected_success_gain_over_base": 0.0,
546
+ "selected_utility": 0.10283424705266953,
547
+ "selected_utility_gain_over_base": 0.0,
548
+ "selector_regret": 0.0,
549
+ "success_selector_gap": 0.0,
550
+ "success_support_gap": 0.0,
551
+ "support_gap": 0.2555490341037512,
552
+ "tau": 0.08716187110665122
553
+ },
554
+ "1461069713/train0": {
555
+ "base_success": 0.0,
556
+ "base_utility": 0.0939030647277832,
557
+ "coverage": 0.0,
558
+ "execute_generated": 0.0,
559
+ "fallback_rate": 1.0,
560
+ "hidden_chart_oracle_success": 0.0,
561
+ "hidden_chart_oracle_utility": 0.1593424528837204,
562
+ "outcome_ptr": 0.0,
563
+ "predicted_margin": 0.04000927889259848,
564
+ "proposal_oracle_success": 0.0,
565
+ "proposal_oracle_utility": 0.06019836664199829,
566
+ "selected_candidate_index": 3.0,
567
+ "selected_success": 0.0,
568
+ "selected_success_gain_over_base": 0.0,
569
+ "selected_utility": 0.0939030647277832,
570
+ "selected_utility_gain_over_base": 0.0,
571
+ "selector_regret": 0.0,
572
+ "success_selector_gap": 0.0,
573
+ "success_support_gap": 0.0,
574
+ "support_gap": 0.0991440862417221,
575
+ "tau": 0.08716187110665122
576
+ },
577
+ "1461069713/train1": {
578
+ "base_success": 0.0,
579
+ "base_utility": 0.0939030647277832,
580
+ "coverage": 1.0,
581
+ "execute_generated": 1.0,
582
+ "fallback_rate": 0.0,
583
+ "hidden_chart_oracle_success": 0.0,
584
+ "hidden_chart_oracle_utility": 0.1593424528837204,
585
+ "outcome_ptr": 0.0,
586
+ "predicted_margin": 0.20548615956297334,
587
+ "proposal_oracle_success": 0.0,
588
+ "proposal_oracle_utility": 0.016585012897849083,
589
+ "selected_candidate_index": 1.0,
590
+ "selected_success": 0.0,
591
+ "selected_success_gain_over_base": 0.0,
592
+ "selected_utility": 0.0066074528731405735,
593
+ "selected_utility_gain_over_base": -0.08729561185464263,
594
+ "selector_regret": 0.00997756002470851,
595
+ "success_selector_gap": 0.0,
596
+ "success_support_gap": 0.0,
597
+ "support_gap": 0.14275743998587132,
598
+ "tau": 0.08716187110665122
599
+ },
600
+ "1461069713/train2": {
601
+ "base_success": 0.0,
602
+ "base_utility": 0.0939030647277832,
603
+ "coverage": 1.0,
604
+ "execute_generated": 1.0,
605
+ "fallback_rate": 0.0,
606
+ "hidden_chart_oracle_success": 0.0,
607
+ "hidden_chart_oracle_utility": 0.1593424528837204,
608
+ "outcome_ptr": 1.0,
609
+ "predicted_margin": 0.128115100020375,
610
+ "proposal_oracle_success": 0.0,
611
+ "proposal_oracle_utility": 0.09970339387655258,
612
+ "selected_candidate_index": 1.0,
613
+ "selected_success": 0.0,
614
+ "selected_success_gain_over_base": 0.0,
615
+ "selected_utility": 0.09168759733438492,
616
+ "selected_utility_gain_over_base": -0.002215467393398285,
617
+ "selector_regret": 0.008015796542167664,
618
+ "success_selector_gap": 0.0,
619
+ "success_support_gap": 0.0,
620
+ "support_gap": 0.059639059007167816,
621
+ "tau": 0.08716187110665122
622
+ },
623
+ "1480856810/train0": {
624
+ "base_success": 0.0,
625
+ "base_utility": 0.08703407645225525,
626
+ "coverage": 1.0,
627
+ "execute_generated": 1.0,
628
+ "fallback_rate": 0.0,
629
+ "hidden_chart_oracle_success": 0.0,
630
+ "hidden_chart_oracle_utility": 0.20948036015033722,
631
+ "outcome_ptr": 1.0,
632
+ "predicted_margin": 0.1780172286211576,
633
+ "proposal_oracle_success": 1.0,
634
+ "proposal_oracle_utility": 2.0,
635
+ "selected_candidate_index": 0.0,
636
+ "selected_success": 1.0,
637
+ "selected_success_gain_over_base": 1.0,
638
+ "selected_utility": 2.0,
639
+ "selected_utility_gain_over_base": 1.9129659235477448,
640
+ "selector_regret": 0.0,
641
+ "success_selector_gap": 0.0,
642
+ "success_support_gap": 0.0,
643
+ "support_gap": 0.0,
644
+ "tau": 0.08716187110665122
645
+ },
646
+ "1480856810/train1": {
647
+ "base_success": 0.0,
648
+ "base_utility": 0.08703407645225525,
649
+ "coverage": 1.0,
650
+ "execute_generated": 1.0,
651
+ "fallback_rate": 0.0,
652
+ "hidden_chart_oracle_success": 0.0,
653
+ "hidden_chart_oracle_utility": 0.20948036015033722,
654
+ "outcome_ptr": 1.0,
655
+ "predicted_margin": 0.42085297639903213,
656
+ "proposal_oracle_success": 1.0,
657
+ "proposal_oracle_utility": 2.0,
658
+ "selected_candidate_index": 6.0,
659
+ "selected_success": 1.0,
660
+ "selected_success_gain_over_base": 1.0,
661
+ "selected_utility": 2.0,
662
+ "selected_utility_gain_over_base": 1.9129659235477448,
663
+ "selector_regret": 0.0,
664
+ "success_selector_gap": 0.0,
665
+ "success_support_gap": 0.0,
666
+ "support_gap": 0.0,
667
+ "tau": 0.08716187110665122
668
+ },
669
+ "1480856810/train2": {
670
+ "base_success": 0.0,
671
+ "base_utility": 0.08703407645225525,
672
+ "coverage": 1.0,
673
+ "execute_generated": 1.0,
674
+ "fallback_rate": 0.0,
675
+ "hidden_chart_oracle_success": 0.0,
676
+ "hidden_chart_oracle_utility": 0.20948036015033722,
677
+ "outcome_ptr": 1.0,
678
+ "predicted_margin": 0.19013993030083898,
679
+ "proposal_oracle_success": 1.0,
680
+ "proposal_oracle_utility": 2.0,
681
+ "selected_candidate_index": 2.0,
682
+ "selected_success": 0.0,
683
+ "selected_success_gain_over_base": 0.0,
684
+ "selected_utility": 0.21894000470638275,
685
+ "selected_utility_gain_over_base": 0.1319059282541275,
686
+ "selector_regret": 1.7810599952936172,
687
+ "success_selector_gap": 1.0,
688
+ "success_support_gap": 0.0,
689
+ "support_gap": 0.0,
690
+ "tau": 0.08716187110665122
691
+ },
692
+ "1509980664/train0": {
693
+ "base_success": 0.0,
694
+ "base_utility": 0.8334100246429443,
695
+ "coverage": 0.0,
696
+ "execute_generated": 0.0,
697
+ "fallback_rate": 1.0,
698
+ "hidden_chart_oracle_success": 1.0,
699
+ "hidden_chart_oracle_utility": 2.0,
700
+ "outcome_ptr": 1.0,
701
+ "predicted_margin": 0.05968870632207461,
702
+ "proposal_oracle_success": 1.0,
703
+ "proposal_oracle_utility": 2.0,
704
+ "selected_candidate_index": 1.0,
705
+ "selected_success": 0.0,
706
+ "selected_success_gain_over_base": 0.0,
707
+ "selected_utility": 0.8334100246429443,
708
+ "selected_utility_gain_over_base": 0.0,
709
+ "selector_regret": 1.1665899753570557,
710
+ "success_selector_gap": 1.0,
711
+ "success_support_gap": 0.0,
712
+ "support_gap": 0.0,
713
+ "tau": 0.08716187110665122
714
+ },
715
+ "1509980664/train1": {
716
+ "base_success": 0.0,
717
+ "base_utility": 0.8334100246429443,
718
+ "coverage": 1.0,
719
+ "execute_generated": 1.0,
720
+ "fallback_rate": 0.0,
721
+ "hidden_chart_oracle_success": 1.0,
722
+ "hidden_chart_oracle_utility": 2.0,
723
+ "outcome_ptr": 1.0,
724
+ "predicted_margin": 0.09344588262600806,
725
+ "proposal_oracle_success": 1.0,
726
+ "proposal_oracle_utility": 2.0,
727
+ "selected_candidate_index": 5.0,
728
+ "selected_success": 0.0,
729
+ "selected_success_gain_over_base": 0.0,
730
+ "selected_utility": 0.8125736117362976,
731
+ "selected_utility_gain_over_base": -0.02083641290664673,
732
+ "selector_regret": 1.1874263882637024,
733
+ "success_selector_gap": 1.0,
734
+ "success_support_gap": 0.0,
735
+ "support_gap": 0.0,
736
+ "tau": 0.08716187110665122
737
+ },
738
+ "1509980664/train2": {
739
+ "base_success": 0.0,
740
+ "base_utility": 0.8334100246429443,
741
+ "coverage": 1.0,
742
+ "execute_generated": 1.0,
743
+ "fallback_rate": 0.0,
744
+ "hidden_chart_oracle_success": 1.0,
745
+ "hidden_chart_oracle_utility": 2.0,
746
+ "outcome_ptr": 0.0,
747
+ "predicted_margin": 0.1474909812049707,
748
+ "proposal_oracle_success": 0.0,
749
+ "proposal_oracle_utility": 0.8162217140197754,
750
+ "selected_candidate_index": 3.0,
751
+ "selected_success": 0.0,
752
+ "selected_success_gain_over_base": 0.0,
753
+ "selected_utility": 0.18445390462875366,
754
+ "selected_utility_gain_over_base": -0.6489561200141907,
755
+ "selector_regret": 0.6317678093910217,
756
+ "success_selector_gap": 0.0,
757
+ "success_support_gap": 1.0,
758
+ "support_gap": 1.1837782859802246,
759
+ "tau": 0.08716187110665122
760
+ },
761
+ "1522272686/train0": {
762
+ "base_success": 0.0,
763
+ "base_utility": 0.5839591026306152,
764
+ "coverage": 0.0,
765
+ "execute_generated": 0.0,
766
+ "fallback_rate": 1.0,
767
+ "hidden_chart_oracle_success": 1.0,
768
+ "hidden_chart_oracle_utility": 2.0,
769
+ "outcome_ptr": 0.0,
770
+ "predicted_margin": -0.11766601647234157,
771
+ "proposal_oracle_success": 0.0,
772
+ "proposal_oracle_utility": 0.18325337767601013,
773
+ "selected_candidate_index": 4.0,
774
+ "selected_success": 0.0,
775
+ "selected_success_gain_over_base": 0.0,
776
+ "selected_utility": 0.5839591026306152,
777
+ "selected_utility_gain_over_base": 0.0,
778
+ "selector_regret": 0.0,
779
+ "success_selector_gap": 0.0,
780
+ "success_support_gap": 1.0,
781
+ "support_gap": 1.8167466223239899,
782
+ "tau": 0.08716187110665122
783
+ },
784
+ "1522272686/train1": {
785
+ "base_success": 0.0,
786
+ "base_utility": 0.5839591026306152,
787
+ "coverage": 0.0,
788
+ "execute_generated": 0.0,
789
+ "fallback_rate": 1.0,
790
+ "hidden_chart_oracle_success": 1.0,
791
+ "hidden_chart_oracle_utility": 2.0,
792
+ "outcome_ptr": 0.0,
793
+ "predicted_margin": -0.030938592468850334,
794
+ "proposal_oracle_success": 0.0,
795
+ "proposal_oracle_utility": 0.07014383375644684,
796
+ "selected_candidate_index": 4.0,
797
+ "selected_success": 0.0,
798
+ "selected_success_gain_over_base": 0.0,
799
+ "selected_utility": 0.5839591026306152,
800
+ "selected_utility_gain_over_base": 0.0,
801
+ "selector_regret": 0.0,
802
+ "success_selector_gap": 0.0,
803
+ "success_support_gap": 1.0,
804
+ "support_gap": 1.9298561662435532,
805
+ "tau": 0.08716187110665122
806
+ },
807
+ "1522272686/train2": {
808
+ "base_success": 0.0,
809
+ "base_utility": 0.5839591026306152,
810
+ "coverage": 1.0,
811
+ "execute_generated": 1.0,
812
+ "fallback_rate": 0.0,
813
+ "hidden_chart_oracle_success": 1.0,
814
+ "hidden_chart_oracle_utility": 2.0,
815
+ "outcome_ptr": 0.0,
816
+ "predicted_margin": 0.09873387353166901,
817
+ "proposal_oracle_success": 0.0,
818
+ "proposal_oracle_utility": 0.14937306940555573,
819
+ "selected_candidate_index": 4.0,
820
+ "selected_success": 0.0,
821
+ "selected_success_gain_over_base": 0.0,
822
+ "selected_utility": 0.10954894870519638,
823
+ "selected_utility_gain_over_base": -0.47441015392541885,
824
+ "selector_regret": 0.039824120700359344,
825
+ "success_selector_gap": 0.0,
826
+ "success_support_gap": 1.0,
827
+ "support_gap": 1.8506269305944443,
828
+ "tau": 0.08716187110665122
829
+ },
830
+ "1549324805/train0": {
831
+ "base_success": 0.0,
832
+ "base_utility": 0.10451483726501465,
833
+ "coverage": 0.0,
834
+ "execute_generated": 0.0,
835
+ "fallback_rate": 1.0,
836
+ "hidden_chart_oracle_success": 0.0,
837
+ "hidden_chart_oracle_utility": 0.2974560856819153,
838
+ "outcome_ptr": 0.0,
839
+ "predicted_margin": 0.062338320403699604,
840
+ "proposal_oracle_success": 0.0,
841
+ "proposal_oracle_utility": 0.0415392704308033,
842
+ "selected_candidate_index": 4.0,
843
+ "selected_success": 0.0,
844
+ "selected_success_gain_over_base": 0.0,
845
+ "selected_utility": 0.10451483726501465,
846
+ "selected_utility_gain_over_base": 0.0,
847
+ "selector_regret": 0.0,
848
+ "success_selector_gap": 0.0,
849
+ "success_support_gap": 0.0,
850
+ "support_gap": 0.255916815251112,
851
+ "tau": 0.08716187110665122
852
+ },
853
+ "1549324805/train1": {
854
+ "base_success": 0.0,
855
+ "base_utility": 0.10451483726501465,
856
+ "coverage": 0.0,
857
+ "execute_generated": 0.0,
858
+ "fallback_rate": 1.0,
859
+ "hidden_chart_oracle_success": 0.0,
860
+ "hidden_chart_oracle_utility": 0.2974560856819153,
861
+ "outcome_ptr": 1.0,
862
+ "predicted_margin": 0.036782065256196234,
863
+ "proposal_oracle_success": 0.0,
864
+ "proposal_oracle_utility": 0.19892151653766632,
865
+ "selected_candidate_index": 0.0,
866
+ "selected_success": 0.0,
867
+ "selected_success_gain_over_base": 0.0,
868
+ "selected_utility": 0.10451483726501465,
869
+ "selected_utility_gain_over_base": 0.0,
870
+ "selector_regret": 0.09440667927265167,
871
+ "success_selector_gap": 0.0,
872
+ "success_support_gap": 0.0,
873
+ "support_gap": 0.09853456914424896,
874
+ "tau": 0.08716187110665122
875
+ },
876
+ "1549324805/train2": {
877
+ "base_success": 0.0,
878
+ "base_utility": 0.10451483726501465,
879
+ "coverage": 1.0,
880
+ "execute_generated": 1.0,
881
+ "fallback_rate": 0.0,
882
+ "hidden_chart_oracle_success": 0.0,
883
+ "hidden_chart_oracle_utility": 0.2974560856819153,
884
+ "outcome_ptr": 0.0,
885
+ "predicted_margin": 0.2580034922169641,
886
+ "proposal_oracle_success": 0.0,
887
+ "proposal_oracle_utility": 0.09092829376459122,
888
+ "selected_candidate_index": 1.0,
889
+ "selected_success": 0.0,
890
+ "selected_success_gain_over_base": 0.0,
891
+ "selected_utility": 0.04690708592534065,
892
+ "selected_utility_gain_over_base": -0.057607751339673996,
893
+ "selector_regret": 0.044021207839250565,
894
+ "success_selector_gap": 0.0,
895
+ "success_support_gap": 0.0,
896
+ "support_gap": 0.20652779191732407,
897
+ "tau": 0.08716187110665122
898
+ },
899
+ "1578990937/train0": {
900
+ "base_success": 0.0,
901
+ "base_utility": 0.82413649559021,
902
+ "coverage": 1.0,
903
+ "execute_generated": 1.0,
904
+ "fallback_rate": 0.0,
905
+ "hidden_chart_oracle_success": 1.0,
906
+ "hidden_chart_oracle_utility": 2.0,
907
+ "outcome_ptr": 1.0,
908
+ "predicted_margin": 0.13461244568720632,
909
+ "proposal_oracle_success": 1.0,
910
+ "proposal_oracle_utility": 2.0,
911
+ "selected_candidate_index": 7.0,
912
+ "selected_success": 0.0,
913
+ "selected_success_gain_over_base": 0.0,
914
+ "selected_utility": 0.16207581758499146,
915
+ "selected_utility_gain_over_base": -0.6620606780052185,
916
+ "selector_regret": 1.8379241824150085,
917
+ "success_selector_gap": 1.0,
918
+ "success_support_gap": 0.0,
919
+ "support_gap": 0.0,
920
+ "tau": 0.08716187110665122
921
+ },
922
+ "1578990937/train1": {
923
+ "base_success": 0.0,
924
+ "base_utility": 0.82413649559021,
925
+ "coverage": 1.0,
926
+ "execute_generated": 1.0,
927
+ "fallback_rate": 0.0,
928
+ "hidden_chart_oracle_success": 1.0,
929
+ "hidden_chart_oracle_utility": 2.0,
930
+ "outcome_ptr": 1.0,
931
+ "predicted_margin": 0.13103425163955518,
932
+ "proposal_oracle_success": 1.0,
933
+ "proposal_oracle_utility": 2.0,
934
+ "selected_candidate_index": 2.0,
935
+ "selected_success": 0.0,
936
+ "selected_success_gain_over_base": 0.0,
937
+ "selected_utility": 0.1305437982082367,
938
+ "selected_utility_gain_over_base": -0.6935926973819733,
939
+ "selector_regret": 1.8694562017917633,
940
+ "success_selector_gap": 1.0,
941
+ "success_support_gap": 0.0,
942
+ "support_gap": 0.0,
943
+ "tau": 0.08716187110665122
944
+ },
945
+ "1578990937/train2": {
946
+ "base_success": 0.0,
947
+ "base_utility": 0.82413649559021,
948
+ "coverage": 1.0,
949
+ "execute_generated": 1.0,
950
+ "fallback_rate": 0.0,
951
+ "hidden_chart_oracle_success": 1.0,
952
+ "hidden_chart_oracle_utility": 2.0,
953
+ "outcome_ptr": 1.0,
954
+ "predicted_margin": 0.16731509907857675,
955
+ "proposal_oracle_success": 1.0,
956
+ "proposal_oracle_utility": 2.0,
957
+ "selected_candidate_index": 2.0,
958
+ "selected_success": 1.0,
959
+ "selected_success_gain_over_base": 1.0,
960
+ "selected_utility": 2.0,
961
+ "selected_utility_gain_over_base": 1.17586350440979,
962
+ "selector_regret": 0.0,
963
+ "success_selector_gap": 0.0,
964
+ "success_support_gap": 0.0,
965
+ "support_gap": 0.0,
966
+ "tau": 0.08716187110665122
967
+ },
968
+ "1747178030/train0": {
969
+ "base_success": 1.0,
970
+ "base_utility": 2.0,
971
+ "coverage": 0.0,
972
+ "execute_generated": 0.0,
973
+ "fallback_rate": 1.0,
974
+ "hidden_chart_oracle_success": 1.0,
975
+ "hidden_chart_oracle_utility": 2.0,
976
+ "outcome_ptr": 0.0,
977
+ "predicted_margin": 0.08446609447311754,
978
+ "proposal_oracle_success": 1.0,
979
+ "proposal_oracle_utility": 2.0,
980
+ "selected_candidate_index": 1.0,
981
+ "selected_success": 1.0,
982
+ "selected_success_gain_over_base": 0.0,
983
+ "selected_utility": 2.0,
984
+ "selected_utility_gain_over_base": 0.0,
985
+ "selector_regret": 0.0,
986
+ "success_selector_gap": 0.0,
987
+ "success_support_gap": 0.0,
988
+ "support_gap": 0.0,
989
+ "tau": 0.08716187110665122
990
+ },
991
+ "1747178030/train1": {
992
+ "base_success": 1.0,
993
+ "base_utility": 2.0,
994
+ "coverage": 1.0,
995
+ "execute_generated": 1.0,
996
+ "fallback_rate": 0.0,
997
+ "hidden_chart_oracle_success": 1.0,
998
+ "hidden_chart_oracle_utility": 2.0,
999
+ "outcome_ptr": 0.0,
1000
+ "predicted_margin": 0.15945400835473367,
1001
+ "proposal_oracle_success": 1.0,
1002
+ "proposal_oracle_utility": 2.0,
1003
+ "selected_candidate_index": 4.0,
1004
+ "selected_success": 1.0,
1005
+ "selected_success_gain_over_base": 0.0,
1006
+ "selected_utility": 2.0,
1007
+ "selected_utility_gain_over_base": 0.0,
1008
+ "selector_regret": 0.0,
1009
+ "success_selector_gap": 0.0,
1010
+ "success_support_gap": 0.0,
1011
+ "support_gap": 0.0,
1012
+ "tau": 0.08716187110665122
1013
+ },
1014
+ "1747178030/train2": {
1015
+ "base_success": 1.0,
1016
+ "base_utility": 2.0,
1017
+ "coverage": 1.0,
1018
+ "execute_generated": 1.0,
1019
+ "fallback_rate": 0.0,
1020
+ "hidden_chart_oracle_success": 1.0,
1021
+ "hidden_chart_oracle_utility": 2.0,
1022
+ "outcome_ptr": 0.0,
1023
+ "predicted_margin": 0.11315031323919739,
1024
+ "proposal_oracle_success": 1.0,
1025
+ "proposal_oracle_utility": 2.0,
1026
+ "selected_candidate_index": 6.0,
1027
+ "selected_success": 1.0,
1028
+ "selected_success_gain_over_base": 0.0,
1029
+ "selected_utility": 2.0,
1030
+ "selected_utility_gain_over_base": 0.0,
1031
+ "selector_regret": 0.0,
1032
+ "success_selector_gap": 0.0,
1033
+ "success_support_gap": 0.0,
1034
+ "support_gap": 0.0,
1035
+ "tau": 0.08716187110665122
1036
+ },
1037
+ "1790188787/train0": {
1038
+ "base_success": 0.0,
1039
+ "base_utility": 0.12476746737957001,
1040
+ "coverage": 1.0,
1041
+ "execute_generated": 1.0,
1042
+ "fallback_rate": 0.0,
1043
+ "hidden_chart_oracle_success": 1.0,
1044
+ "hidden_chart_oracle_utility": 2.0,
1045
+ "outcome_ptr": 0.0,
1046
+ "predicted_margin": 0.1467429189047256,
1047
+ "proposal_oracle_success": 0.0,
1048
+ "proposal_oracle_utility": 0.09791924059391022,
1049
+ "selected_candidate_index": 4.0,
1050
+ "selected_success": 0.0,
1051
+ "selected_success_gain_over_base": 0.0,
1052
+ "selected_utility": 0.06609509885311127,
1053
+ "selected_utility_gain_over_base": -0.05867236852645874,
1054
+ "selector_regret": 0.03182414174079895,
1055
+ "success_selector_gap": 0.0,
1056
+ "success_support_gap": 1.0,
1057
+ "support_gap": 1.9020807594060898,
1058
+ "tau": 0.08716187110665122
1059
+ },
1060
+ "1790188787/train1": {
1061
+ "base_success": 0.0,
1062
+ "base_utility": 0.12476746737957001,
1063
+ "coverage": 1.0,
1064
+ "execute_generated": 1.0,
1065
+ "fallback_rate": 0.0,
1066
+ "hidden_chart_oracle_success": 1.0,
1067
+ "hidden_chart_oracle_utility": 2.0,
1068
+ "outcome_ptr": 1.0,
1069
+ "predicted_margin": 0.09406763885688091,
1070
+ "proposal_oracle_success": 0.0,
1071
+ "proposal_oracle_utility": 0.13031470775604248,
1072
+ "selected_candidate_index": 7.0,
1073
+ "selected_success": 0.0,
1074
+ "selected_success_gain_over_base": 0.0,
1075
+ "selected_utility": 0.0009841322898864746,
1076
+ "selected_utility_gain_over_base": -0.12378333508968353,
1077
+ "selector_regret": 0.129330575466156,
1078
+ "success_selector_gap": 0.0,
1079
+ "success_support_gap": 1.0,
1080
+ "support_gap": 1.8696852922439575,
1081
+ "tau": 0.08716187110665122
1082
+ },
1083
+ "1790188787/train2": {
1084
+ "base_success": 0.0,
1085
+ "base_utility": 0.12476746737957001,
1086
+ "coverage": 1.0,
1087
+ "execute_generated": 1.0,
1088
+ "fallback_rate": 0.0,
1089
+ "hidden_chart_oracle_success": 1.0,
1090
+ "hidden_chart_oracle_utility": 2.0,
1091
+ "outcome_ptr": 1.0,
1092
+ "predicted_margin": 0.15565215755253797,
1093
+ "proposal_oracle_success": 1.0,
1094
+ "proposal_oracle_utility": 2.0,
1095
+ "selected_candidate_index": 1.0,
1096
+ "selected_success": 0.0,
1097
+ "selected_success_gain_over_base": 0.0,
1098
+ "selected_utility": 0.1341685652732849,
1099
+ "selected_utility_gain_over_base": 0.009401097893714905,
1100
+ "selector_regret": 1.865831434726715,
1101
+ "success_selector_gap": 1.0,
1102
+ "success_support_gap": 0.0,
1103
+ "support_gap": 0.0,
1104
+ "tau": 0.08716187110665122
1105
+ },
1106
+ "1822210471/train0": {
1107
+ "base_success": 0.0,
1108
+ "base_utility": 0.07191530615091324,
1109
+ "coverage": 0.0,
1110
+ "execute_generated": 0.0,
1111
+ "fallback_rate": 1.0,
1112
+ "hidden_chart_oracle_success": 1.0,
1113
+ "hidden_chart_oracle_utility": 2.0,
1114
+ "outcome_ptr": 1.0,
1115
+ "predicted_margin": -0.004889711129218405,
1116
+ "proposal_oracle_success": 1.0,
1117
+ "proposal_oracle_utility": 2.0,
1118
+ "selected_candidate_index": 3.0,
1119
+ "selected_success": 0.0,
1120
+ "selected_success_gain_over_base": 0.0,
1121
+ "selected_utility": 0.07191530615091324,
1122
+ "selected_utility_gain_over_base": 0.0,
1123
+ "selector_regret": 1.9280846938490868,
1124
+ "success_selector_gap": 1.0,
1125
+ "success_support_gap": 0.0,
1126
+ "support_gap": 0.0,
1127
+ "tau": 0.08716187110665122
1128
+ },
1129
+ "1822210471/train1": {
1130
+ "base_success": 0.0,
1131
+ "base_utility": 0.07191530615091324,
1132
+ "coverage": 0.0,
1133
+ "execute_generated": 0.0,
1134
+ "fallback_rate": 1.0,
1135
+ "hidden_chart_oracle_success": 1.0,
1136
+ "hidden_chart_oracle_utility": 2.0,
1137
+ "outcome_ptr": 1.0,
1138
+ "predicted_margin": 0.01936054008091473,
1139
+ "proposal_oracle_success": 1.0,
1140
+ "proposal_oracle_utility": 2.0,
1141
+ "selected_candidate_index": 3.0,
1142
+ "selected_success": 0.0,
1143
+ "selected_success_gain_over_base": 0.0,
1144
+ "selected_utility": 0.07191530615091324,
1145
+ "selected_utility_gain_over_base": 0.0,
1146
+ "selector_regret": 1.9280846938490868,
1147
+ "success_selector_gap": 1.0,
1148
+ "success_support_gap": 0.0,
1149
+ "support_gap": 0.0,
1150
+ "tau": 0.08716187110665122
1151
+ },
1152
+ "1822210471/train2": {
1153
+ "base_success": 0.0,
1154
+ "base_utility": 0.07191530615091324,
1155
+ "coverage": 0.0,
1156
+ "execute_generated": 0.0,
1157
+ "fallback_rate": 1.0,
1158
+ "hidden_chart_oracle_success": 1.0,
1159
+ "hidden_chart_oracle_utility": 2.0,
1160
+ "outcome_ptr": 1.0,
1161
+ "predicted_margin": -0.012742203632338828,
1162
+ "proposal_oracle_success": 1.0,
1163
+ "proposal_oracle_utility": 2.0,
1164
+ "selected_candidate_index": 7.0,
1165
+ "selected_success": 0.0,
1166
+ "selected_success_gain_over_base": 0.0,
1167
+ "selected_utility": 0.07191530615091324,
1168
+ "selected_utility_gain_over_base": 0.0,
1169
+ "selector_regret": 1.9280846938490868,
1170
+ "success_selector_gap": 1.0,
1171
+ "success_support_gap": 0.0,
1172
+ "support_gap": 0.0,
1173
+ "tau": 0.08716187110665122
1174
+ },
1175
+ "1845489013/train0": {
1176
+ "base_success": 1.0,
1177
+ "base_utility": 2.0,
1178
+ "coverage": 0.0,
1179
+ "execute_generated": 0.0,
1180
+ "fallback_rate": 1.0,
1181
+ "hidden_chart_oracle_success": 1.0,
1182
+ "hidden_chart_oracle_utility": 2.0,
1183
+ "outcome_ptr": 0.0,
1184
+ "predicted_margin": 0.048965368633265796,
1185
+ "proposal_oracle_success": 1.0,
1186
+ "proposal_oracle_utility": 2.0,
1187
+ "selected_candidate_index": 6.0,
1188
+ "selected_success": 1.0,
1189
+ "selected_success_gain_over_base": 0.0,
1190
+ "selected_utility": 2.0,
1191
+ "selected_utility_gain_over_base": 0.0,
1192
+ "selector_regret": 0.0,
1193
+ "success_selector_gap": 0.0,
1194
+ "success_support_gap": 0.0,
1195
+ "support_gap": 0.0,
1196
+ "tau": 0.08716187110665122
1197
+ },
1198
+ "1845489013/train1": {
1199
+ "base_success": 1.0,
1200
+ "base_utility": 2.0,
1201
+ "coverage": 0.0,
1202
+ "execute_generated": 0.0,
1203
+ "fallback_rate": 1.0,
1204
+ "hidden_chart_oracle_success": 1.0,
1205
+ "hidden_chart_oracle_utility": 2.0,
1206
+ "outcome_ptr": 0.0,
1207
+ "predicted_margin": 0.0366164295699445,
1208
+ "proposal_oracle_success": 1.0,
1209
+ "proposal_oracle_utility": 2.0,
1210
+ "selected_candidate_index": 3.0,
1211
+ "selected_success": 1.0,
1212
+ "selected_success_gain_over_base": 0.0,
1213
+ "selected_utility": 2.0,
1214
+ "selected_utility_gain_over_base": 0.0,
1215
+ "selector_regret": 0.0,
1216
+ "success_selector_gap": 0.0,
1217
+ "success_support_gap": 0.0,
1218
+ "support_gap": 0.0,
1219
+ "tau": 0.08716187110665122
1220
+ },
1221
+ "1845489013/train2": {
1222
+ "base_success": 1.0,
1223
+ "base_utility": 2.0,
1224
+ "coverage": 0.0,
1225
+ "execute_generated": 0.0,
1226
+ "fallback_rate": 1.0,
1227
+ "hidden_chart_oracle_success": 1.0,
1228
+ "hidden_chart_oracle_utility": 2.0,
1229
+ "outcome_ptr": 0.0,
1230
+ "predicted_margin": 0.037789214067942875,
1231
+ "proposal_oracle_success": 1.0,
1232
+ "proposal_oracle_utility": 2.0,
1233
+ "selected_candidate_index": 6.0,
1234
+ "selected_success": 1.0,
1235
+ "selected_success_gain_over_base": 0.0,
1236
+ "selected_utility": 2.0,
1237
+ "selected_utility_gain_over_base": 0.0,
1238
+ "selector_regret": 0.0,
1239
+ "success_selector_gap": 0.0,
1240
+ "success_support_gap": 0.0,
1241
+ "support_gap": 0.0,
1242
+ "tau": 0.08716187110665122
1243
+ },
1244
+ "1999022574/train0": {
1245
+ "base_success": 1.0,
1246
+ "base_utility": 2.0,
1247
+ "coverage": 0.0,
1248
+ "execute_generated": 0.0,
1249
+ "fallback_rate": 1.0,
1250
+ "hidden_chart_oracle_success": 1.0,
1251
+ "hidden_chart_oracle_utility": 2.0,
1252
+ "outcome_ptr": 0.0,
1253
+ "predicted_margin": 0.04875165751509537,
1254
+ "proposal_oracle_success": 1.0,
1255
+ "proposal_oracle_utility": 2.0,
1256
+ "selected_candidate_index": 2.0,
1257
+ "selected_success": 1.0,
1258
+ "selected_success_gain_over_base": 0.0,
1259
+ "selected_utility": 2.0,
1260
+ "selected_utility_gain_over_base": 0.0,
1261
+ "selector_regret": 0.0,
1262
+ "success_selector_gap": 0.0,
1263
+ "success_support_gap": 0.0,
1264
+ "support_gap": 0.0,
1265
+ "tau": 0.08716187110665122
1266
+ },
1267
+ "1999022574/train1": {
1268
+ "base_success": 1.0,
1269
+ "base_utility": 2.0,
1270
+ "coverage": 0.0,
1271
+ "execute_generated": 0.0,
1272
+ "fallback_rate": 1.0,
1273
+ "hidden_chart_oracle_success": 1.0,
1274
+ "hidden_chart_oracle_utility": 2.0,
1275
+ "outcome_ptr": 0.0,
1276
+ "predicted_margin": 0.08112579475085248,
1277
+ "proposal_oracle_success": 1.0,
1278
+ "proposal_oracle_utility": 2.0,
1279
+ "selected_candidate_index": 4.0,
1280
+ "selected_success": 1.0,
1281
+ "selected_success_gain_over_base": 0.0,
1282
+ "selected_utility": 2.0,
1283
+ "selected_utility_gain_over_base": 0.0,
1284
+ "selector_regret": 0.0,
1285
+ "success_selector_gap": 0.0,
1286
+ "success_support_gap": 0.0,
1287
+ "support_gap": 0.0,
1288
+ "tau": 0.08716187110665122
1289
+ },
1290
+ "1999022574/train2": {
1291
+ "base_success": 1.0,
1292
+ "base_utility": 2.0,
1293
+ "coverage": 1.0,
1294
+ "execute_generated": 1.0,
1295
+ "fallback_rate": 0.0,
1296
+ "hidden_chart_oracle_success": 1.0,
1297
+ "hidden_chart_oracle_utility": 2.0,
1298
+ "outcome_ptr": 0.0,
1299
+ "predicted_margin": 0.09966294159986532,
1300
+ "proposal_oracle_success": 1.0,
1301
+ "proposal_oracle_utility": 2.0,
1302
+ "selected_candidate_index": 4.0,
1303
+ "selected_success": 1.0,
1304
+ "selected_success_gain_over_base": 0.0,
1305
+ "selected_utility": 2.0,
1306
+ "selected_utility_gain_over_base": 0.0,
1307
+ "selector_regret": 0.0,
1308
+ "success_selector_gap": 0.0,
1309
+ "success_support_gap": 0.0,
1310
+ "support_gap": 0.0,
1311
+ "tau": 0.08716187110665122
1312
+ },
1313
+ "2012211287/train0": {
1314
+ "base_success": 0.0,
1315
+ "base_utility": 0.7236053347587585,
1316
+ "coverage": 0.0,
1317
+ "execute_generated": 0.0,
1318
+ "fallback_rate": 1.0,
1319
+ "hidden_chart_oracle_success": 1.0,
1320
+ "hidden_chart_oracle_utility": 2.0,
1321
+ "outcome_ptr": 1.0,
1322
+ "predicted_margin": 0.025740406212082005,
1323
+ "proposal_oracle_success": 1.0,
1324
+ "proposal_oracle_utility": 2.0,
1325
+ "selected_candidate_index": 5.0,
1326
+ "selected_success": 0.0,
1327
+ "selected_success_gain_over_base": 0.0,
1328
+ "selected_utility": 0.7236053347587585,
1329
+ "selected_utility_gain_over_base": 0.0,
1330
+ "selector_regret": 1.2763946652412415,
1331
+ "success_selector_gap": 1.0,
1332
+ "success_support_gap": 0.0,
1333
+ "support_gap": 0.0,
1334
+ "tau": 0.08716187110665122
1335
+ },
1336
+ "2012211287/train1": {
1337
+ "base_success": 0.0,
1338
+ "base_utility": 0.7236053347587585,
1339
+ "coverage": 0.0,
1340
+ "execute_generated": 0.0,
1341
+ "fallback_rate": 1.0,
1342
+ "hidden_chart_oracle_success": 1.0,
1343
+ "hidden_chart_oracle_utility": 2.0,
1344
+ "outcome_ptr": 1.0,
1345
+ "predicted_margin": -0.0387201991331603,
1346
+ "proposal_oracle_success": 1.0,
1347
+ "proposal_oracle_utility": 2.0,
1348
+ "selected_candidate_index": 7.0,
1349
+ "selected_success": 0.0,
1350
+ "selected_success_gain_over_base": 0.0,
1351
+ "selected_utility": 0.7236053347587585,
1352
+ "selected_utility_gain_over_base": 0.0,
1353
+ "selector_regret": 1.2763946652412415,
1354
+ "success_selector_gap": 1.0,
1355
+ "success_support_gap": 0.0,
1356
+ "support_gap": 0.0,
1357
+ "tau": 0.08716187110665122
1358
+ },
1359
+ "2012211287/train2": {
1360
+ "base_success": 0.0,
1361
+ "base_utility": 0.7236053347587585,
1362
+ "coverage": 0.0,
1363
+ "execute_generated": 0.0,
1364
+ "fallback_rate": 1.0,
1365
+ "hidden_chart_oracle_success": 1.0,
1366
+ "hidden_chart_oracle_utility": 2.0,
1367
+ "outcome_ptr": 1.0,
1368
+ "predicted_margin": -0.07824933821700998,
1369
+ "proposal_oracle_success": 1.0,
1370
+ "proposal_oracle_utility": 2.0,
1371
+ "selected_candidate_index": 6.0,
1372
+ "selected_success": 0.0,
1373
+ "selected_success_gain_over_base": 0.0,
1374
+ "selected_utility": 0.7236053347587585,
1375
+ "selected_utility_gain_over_base": 0.0,
1376
+ "selector_regret": 1.2763946652412415,
1377
+ "success_selector_gap": 1.0,
1378
+ "success_support_gap": 0.0,
1379
+ "support_gap": 0.0,
1380
+ "tau": 0.08716187110665122
1381
+ },
1382
+ "2038594191/train0": {
1383
+ "base_success": 1.0,
1384
+ "base_utility": 2.0,
1385
+ "coverage": 0.0,
1386
+ "execute_generated": 0.0,
1387
+ "fallback_rate": 1.0,
1388
+ "hidden_chart_oracle_success": 1.0,
1389
+ "hidden_chart_oracle_utility": 2.0,
1390
+ "outcome_ptr": 0.0,
1391
+ "predicted_margin": -0.042660809330536634,
1392
+ "proposal_oracle_success": 1.0,
1393
+ "proposal_oracle_utility": 2.0,
1394
+ "selected_candidate_index": 3.0,
1395
+ "selected_success": 1.0,
1396
+ "selected_success_gain_over_base": 0.0,
1397
+ "selected_utility": 2.0,
1398
+ "selected_utility_gain_over_base": 0.0,
1399
+ "selector_regret": 0.0,
1400
+ "success_selector_gap": 0.0,
1401
+ "success_support_gap": 0.0,
1402
+ "support_gap": 0.0,
1403
+ "tau": 0.08716187110665122
1404
+ },
1405
+ "2038594191/train1": {
1406
+ "base_success": 1.0,
1407
+ "base_utility": 2.0,
1408
+ "coverage": 0.0,
1409
+ "execute_generated": 0.0,
1410
+ "fallback_rate": 1.0,
1411
+ "hidden_chart_oracle_success": 1.0,
1412
+ "hidden_chart_oracle_utility": 2.0,
1413
+ "outcome_ptr": 0.0,
1414
+ "predicted_margin": -0.012759513540652648,
1415
+ "proposal_oracle_success": 1.0,
1416
+ "proposal_oracle_utility": 2.0,
1417
+ "selected_candidate_index": 0.0,
1418
+ "selected_success": 1.0,
1419
+ "selected_success_gain_over_base": 0.0,
1420
+ "selected_utility": 2.0,
1421
+ "selected_utility_gain_over_base": 0.0,
1422
+ "selector_regret": 0.0,
1423
+ "success_selector_gap": 0.0,
1424
+ "success_support_gap": 0.0,
1425
+ "support_gap": 0.0,
1426
+ "tau": 0.08716187110665122
1427
+ },
1428
+ "2038594191/train2": {
1429
+ "base_success": 1.0,
1430
+ "base_utility": 2.0,
1431
+ "coverage": 0.0,
1432
+ "execute_generated": 0.0,
1433
+ "fallback_rate": 1.0,
1434
+ "hidden_chart_oracle_success": 1.0,
1435
+ "hidden_chart_oracle_utility": 2.0,
1436
+ "outcome_ptr": 0.0,
1437
+ "predicted_margin": -0.031459122278913496,
1438
+ "proposal_oracle_success": 1.0,
1439
+ "proposal_oracle_utility": 2.0,
1440
+ "selected_candidate_index": 1.0,
1441
+ "selected_success": 1.0,
1442
+ "selected_success_gain_over_base": 0.0,
1443
+ "selected_utility": 2.0,
1444
+ "selected_utility_gain_over_base": 0.0,
1445
+ "selector_regret": 0.0,
1446
+ "success_selector_gap": 0.0,
1447
+ "success_support_gap": 0.0,
1448
+ "support_gap": 0.0,
1449
+ "tau": 0.08716187110665122
1450
+ },
1451
+ "2054899126/train0": {
1452
+ "base_success": 1.0,
1453
+ "base_utility": 2.0,
1454
+ "coverage": 0.5,
1455
+ "execute_generated": 0.5,
1456
+ "fallback_rate": 0.5,
1457
+ "hidden_chart_oracle_success": 1.0,
1458
+ "hidden_chart_oracle_utility": 2.0,
1459
+ "outcome_ptr": 0.0,
1460
+ "predicted_margin": 0.019084174365363485,
1461
+ "proposal_oracle_success": 0.5,
1462
+ "proposal_oracle_utility": 1.077283151447773,
1463
+ "selected_candidate_index": 4.5,
1464
+ "selected_success": 1.0,
1465
+ "selected_success_gain_over_base": 0.0,
1466
+ "selected_utility": 2.0,
1467
+ "selected_utility_gain_over_base": 0.0,
1468
+ "selector_regret": 0.0,
1469
+ "success_selector_gap": 0.0,
1470
+ "success_support_gap": 0.5,
1471
+ "support_gap": 0.922716848552227,
1472
+ "tau": 0.08716187110665122
1473
+ },
1474
+ "2054899126/train1": {
1475
+ "base_success": 1.0,
1476
+ "base_utility": 2.0,
1477
+ "coverage": 0.5,
1478
+ "execute_generated": 0.5,
1479
+ "fallback_rate": 0.5,
1480
+ "hidden_chart_oracle_success": 1.0,
1481
+ "hidden_chart_oracle_utility": 2.0,
1482
+ "outcome_ptr": 0.0,
1483
+ "predicted_margin": 0.07453846963513679,
1484
+ "proposal_oracle_success": 0.5,
1485
+ "proposal_oracle_utility": 1.0696808770298958,
1486
+ "selected_candidate_index": 3.0,
1487
+ "selected_success": 1.0,
1488
+ "selected_success_gain_over_base": 0.0,
1489
+ "selected_utility": 2.0,
1490
+ "selected_utility_gain_over_base": 0.0,
1491
+ "selector_regret": 0.0,
1492
+ "success_selector_gap": 0.0,
1493
+ "success_support_gap": 0.5,
1494
+ "support_gap": 0.9303191229701042,
1495
+ "tau": 0.08716187110665122
1496
+ },
1497
+ "2054899126/train2": {
1498
+ "base_success": 1.0,
1499
+ "base_utility": 2.0,
1500
+ "coverage": 0.5,
1501
+ "execute_generated": 0.5,
1502
+ "fallback_rate": 0.5,
1503
+ "hidden_chart_oracle_success": 1.0,
1504
+ "hidden_chart_oracle_utility": 2.0,
1505
+ "outcome_ptr": 0.0,
1506
+ "predicted_margin": -0.06569178303644396,
1507
+ "proposal_oracle_success": 0.5,
1508
+ "proposal_oracle_utility": 1.0455262660980225,
1509
+ "selected_candidate_index": 0.5,
1510
+ "selected_success": 1.0,
1511
+ "selected_success_gain_over_base": 0.0,
1512
+ "selected_utility": 2.0,
1513
+ "selected_utility_gain_over_base": 0.0,
1514
+ "selector_regret": 0.0,
1515
+ "success_selector_gap": 0.0,
1516
+ "success_support_gap": 0.5,
1517
+ "support_gap": 0.9544737339019775,
1518
+ "tau": 0.08716187110665122
1519
+ },
1520
+ "2061635888/train0": {
1521
+ "base_success": 0.0,
1522
+ "base_utility": 0.05225050449371338,
1523
+ "coverage": 0.0,
1524
+ "execute_generated": 0.0,
1525
+ "fallback_rate": 1.0,
1526
+ "hidden_chart_oracle_success": 1.0,
1527
+ "hidden_chart_oracle_utility": 2.0,
1528
+ "outcome_ptr": 1.0,
1529
+ "predicted_margin": 0.006324322195940341,
1530
+ "proposal_oracle_success": 0.0,
1531
+ "proposal_oracle_utility": 0.09799930453300476,
1532
+ "selected_candidate_index": 2.0,
1533
+ "selected_success": 0.0,
1534
+ "selected_success_gain_over_base": 0.0,
1535
+ "selected_utility": 0.05225050449371338,
1536
+ "selected_utility_gain_over_base": 0.0,
1537
+ "selector_regret": 0.04574880003929138,
1538
+ "success_selector_gap": 0.0,
1539
+ "success_support_gap": 1.0,
1540
+ "support_gap": 1.9020006954669952,
1541
+ "tau": 0.08716187110665122
1542
+ },
1543
+ "2061635888/train1": {
1544
+ "base_success": 0.0,
1545
+ "base_utility": 0.05225050449371338,
1546
+ "coverage": 0.0,
1547
+ "execute_generated": 0.0,
1548
+ "fallback_rate": 1.0,
1549
+ "hidden_chart_oracle_success": 1.0,
1550
+ "hidden_chart_oracle_utility": 2.0,
1551
+ "outcome_ptr": 1.0,
1552
+ "predicted_margin": 0.06837147342268665,
1553
+ "proposal_oracle_success": 0.0,
1554
+ "proposal_oracle_utility": 0.07580336928367615,
1555
+ "selected_candidate_index": 1.0,
1556
+ "selected_success": 0.0,
1557
+ "selected_success_gain_over_base": 0.0,
1558
+ "selected_utility": 0.05225050449371338,
1559
+ "selected_utility_gain_over_base": 0.0,
1560
+ "selector_regret": 0.02355286478996277,
1561
+ "success_selector_gap": 0.0,
1562
+ "success_support_gap": 1.0,
1563
+ "support_gap": 1.9241966307163239,
1564
+ "tau": 0.08716187110665122
1565
+ },
1566
+ "2061635888/train2": {
1567
+ "base_success": 0.0,
1568
+ "base_utility": 0.05225050449371338,
1569
+ "coverage": 0.0,
1570
+ "execute_generated": 0.0,
1571
+ "fallback_rate": 1.0,
1572
+ "hidden_chart_oracle_success": 1.0,
1573
+ "hidden_chart_oracle_utility": 2.0,
1574
+ "outcome_ptr": 1.0,
1575
+ "predicted_margin": 0.05341970067966316,
1576
+ "proposal_oracle_success": 0.0,
1577
+ "proposal_oracle_utility": 0.11105132102966309,
1578
+ "selected_candidate_index": 6.0,
1579
+ "selected_success": 0.0,
1580
+ "selected_success_gain_over_base": 0.0,
1581
+ "selected_utility": 0.05225050449371338,
1582
+ "selected_utility_gain_over_base": 0.0,
1583
+ "selector_regret": 0.05880081653594971,
1584
+ "success_selector_gap": 0.0,
1585
+ "success_support_gap": 1.0,
1586
+ "support_gap": 1.888948678970337,
1587
+ "tau": 0.08716187110665122
1588
+ },
1589
+ "211245842/train0": {
1590
+ "base_success": 0.0,
1591
+ "base_utility": 0.19090841710567474,
1592
+ "coverage": 0.0,
1593
+ "execute_generated": 0.0,
1594
+ "fallback_rate": 1.0,
1595
+ "hidden_chart_oracle_success": 1.0,
1596
+ "hidden_chart_oracle_utility": 2.0,
1597
+ "outcome_ptr": 1.0,
1598
+ "predicted_margin": 0.06821062372976947,
1599
+ "proposal_oracle_success": 1.0,
1600
+ "proposal_oracle_utility": 2.0,
1601
+ "selected_candidate_index": 3.0,
1602
+ "selected_success": 0.0,
1603
+ "selected_success_gain_over_base": 0.0,
1604
+ "selected_utility": 0.19090841710567474,
1605
+ "selected_utility_gain_over_base": 0.0,
1606
+ "selector_regret": 1.8090915828943253,
1607
+ "success_selector_gap": 1.0,
1608
+ "success_support_gap": 0.0,
1609
+ "support_gap": 0.0,
1610
+ "tau": 0.08716187110665122
1611
+ },
1612
+ "211245842/train1": {
1613
+ "base_success": 0.0,
1614
+ "base_utility": 0.19090841710567474,
1615
+ "coverage": 1.0,
1616
+ "execute_generated": 1.0,
1617
+ "fallback_rate": 0.0,
1618
+ "hidden_chart_oracle_success": 1.0,
1619
+ "hidden_chart_oracle_utility": 2.0,
1620
+ "outcome_ptr": 1.0,
1621
+ "predicted_margin": 0.12389623305012636,
1622
+ "proposal_oracle_success": 0.0,
1623
+ "proposal_oracle_utility": 0.21295717358589172,
1624
+ "selected_candidate_index": 6.0,
1625
+ "selected_success": 0.0,
1626
+ "selected_success_gain_over_base": 0.0,
1627
+ "selected_utility": 0.1710544228553772,
1628
+ "selected_utility_gain_over_base": -0.019853994250297546,
1629
+ "selector_regret": 0.041902750730514526,
1630
+ "success_selector_gap": 0.0,
1631
+ "success_support_gap": 1.0,
1632
+ "support_gap": 1.7870428264141083,
1633
+ "tau": 0.08716187110665122
1634
+ },
1635
+ "211245842/train2": {
1636
+ "base_success": 0.0,
1637
+ "base_utility": 0.19090841710567474,
1638
+ "coverage": 0.0,
1639
+ "execute_generated": 0.0,
1640
+ "fallback_rate": 1.0,
1641
+ "hidden_chart_oracle_success": 1.0,
1642
+ "hidden_chart_oracle_utility": 2.0,
1643
+ "outcome_ptr": 1.0,
1644
+ "predicted_margin": 0.07939006946358719,
1645
+ "proposal_oracle_success": 0.0,
1646
+ "proposal_oracle_utility": 0.21523693203926086,
1647
+ "selected_candidate_index": 0.0,
1648
+ "selected_success": 0.0,
1649
+ "selected_success_gain_over_base": 0.0,
1650
+ "selected_utility": 0.19090841710567474,
1651
+ "selected_utility_gain_over_base": 0.0,
1652
+ "selector_regret": 0.02432851493358612,
1653
+ "success_selector_gap": 0.0,
1654
+ "success_support_gap": 1.0,
1655
+ "support_gap": 1.7847630679607391,
1656
+ "tau": 0.08716187110665122
1657
+ },
1658
+ "278008524/train0": {
1659
+ "base_success": 0.0,
1660
+ "base_utility": 0.042734989896416664,
1661
+ "coverage": 0.0,
1662
+ "execute_generated": 0.0,
1663
+ "fallback_rate": 1.0,
1664
+ "hidden_chart_oracle_success": 0.5,
1665
+ "hidden_chart_oracle_utility": 1.1139086410403252,
1666
+ "outcome_ptr": 1.0,
1667
+ "predicted_margin": -0.06789446106618917,
1668
+ "proposal_oracle_success": 0.5,
1669
+ "proposal_oracle_utility": 1.0287093725055456,
1670
+ "selected_candidate_index": 4.5,
1671
+ "selected_success": 0.0,
1672
+ "selected_success_gain_over_base": 0.0,
1673
+ "selected_utility": 0.042734989896416664,
1674
+ "selected_utility_gain_over_base": 0.0,
1675
+ "selector_regret": 0.985974382609129,
1676
+ "success_selector_gap": 0.5,
1677
+ "success_support_gap": 0.0,
1678
+ "support_gap": 0.08519926853477955,
1679
+ "tau": 0.08716187110665122
1680
+ },
1681
+ "278008524/train1": {
1682
+ "base_success": 0.0,
1683
+ "base_utility": 0.042734989896416664,
1684
+ "coverage": 0.5,
1685
+ "execute_generated": 0.5,
1686
+ "fallback_rate": 0.5,
1687
+ "hidden_chart_oracle_success": 0.5,
1688
+ "hidden_chart_oracle_utility": 1.1139086410403252,
1689
+ "outcome_ptr": 1.0,
1690
+ "predicted_margin": 0.009910582975323233,
1691
+ "proposal_oracle_success": 0.5,
1692
+ "proposal_oracle_utility": 1.0376745946705341,
1693
+ "selected_candidate_index": 3.5,
1694
+ "selected_success": 0.0,
1695
+ "selected_success_gain_over_base": 0.0,
1696
+ "selected_utility": 0.05399738810956478,
1697
+ "selected_utility_gain_over_base": 0.011262398213148117,
1698
+ "selector_regret": 0.9836772065609694,
1699
+ "success_selector_gap": 0.5,
1700
+ "success_support_gap": 0.0,
1701
+ "support_gap": 0.07623404636979103,
1702
+ "tau": 0.08716187110665122
1703
+ },
1704
+ "278008524/train2": {
1705
+ "base_success": 0.0,
1706
+ "base_utility": 0.042734989896416664,
1707
+ "coverage": 0.0,
1708
+ "execute_generated": 0.0,
1709
+ "fallback_rate": 1.0,
1710
+ "hidden_chart_oracle_success": 0.5,
1711
+ "hidden_chart_oracle_utility": 1.1139086410403252,
1712
+ "outcome_ptr": 1.0,
1713
+ "predicted_margin": -0.003736171272467028,
1714
+ "proposal_oracle_success": 0.5,
1715
+ "proposal_oracle_utility": 1.1359489113092422,
1716
+ "selected_candidate_index": 3.5,
1717
+ "selected_success": 0.0,
1718
+ "selected_success_gain_over_base": 0.0,
1719
+ "selected_utility": 0.042734989896416664,
1720
+ "selected_utility_gain_over_base": 0.0,
1721
+ "selector_regret": 1.0932139214128256,
1722
+ "success_selector_gap": 0.5,
1723
+ "success_support_gap": 0.0,
1724
+ "support_gap": 0.0,
1725
+ "tau": 0.08716187110665122
1726
+ },
1727
+ "307068879/train0": {
1728
+ "base_success": 0.0,
1729
+ "base_utility": 0.6000905632972717,
1730
+ "coverage": 1.0,
1731
+ "execute_generated": 1.0,
1732
+ "fallback_rate": 0.0,
1733
+ "hidden_chart_oracle_success": 1.0,
1734
+ "hidden_chart_oracle_utility": 2.0,
1735
+ "outcome_ptr": 0.0,
1736
+ "predicted_margin": 0.09322755742104462,
1737
+ "proposal_oracle_success": 0.0,
1738
+ "proposal_oracle_utility": 0.2115575522184372,
1739
+ "selected_candidate_index": 3.0,
1740
+ "selected_success": 0.0,
1741
+ "selected_success_gain_over_base": 0.0,
1742
+ "selected_utility": 0.20163558423519135,
1743
+ "selected_utility_gain_over_base": -0.3984549790620804,
1744
+ "selector_regret": 0.00992196798324585,
1745
+ "success_selector_gap": 0.0,
1746
+ "success_support_gap": 1.0,
1747
+ "support_gap": 1.7884424477815628,
1748
+ "tau": 0.08716187110665122
1749
+ },
1750
+ "307068879/train1": {
1751
+ "base_success": 0.0,
1752
+ "base_utility": 0.6000905632972717,
1753
+ "coverage": 0.0,
1754
+ "execute_generated": 0.0,
1755
+ "fallback_rate": 1.0,
1756
+ "hidden_chart_oracle_success": 1.0,
1757
+ "hidden_chart_oracle_utility": 2.0,
1758
+ "outcome_ptr": 0.0,
1759
+ "predicted_margin": -0.042372730661597094,
1760
+ "proposal_oracle_success": 0.0,
1761
+ "proposal_oracle_utility": 0.5876073241233826,
1762
+ "selected_candidate_index": 5.0,
1763
+ "selected_success": 0.0,
1764
+ "selected_success_gain_over_base": 0.0,
1765
+ "selected_utility": 0.6000905632972717,
1766
+ "selected_utility_gain_over_base": 0.0,
1767
+ "selector_regret": 0.0,
1768
+ "success_selector_gap": 0.0,
1769
+ "success_support_gap": 1.0,
1770
+ "support_gap": 1.4123926758766174,
1771
+ "tau": 0.08716187110665122
1772
+ },
1773
+ "307068879/train2": {
1774
+ "base_success": 0.0,
1775
+ "base_utility": 0.6000905632972717,
1776
+ "coverage": 0.0,
1777
+ "execute_generated": 0.0,
1778
+ "fallback_rate": 1.0,
1779
+ "hidden_chart_oracle_success": 1.0,
1780
+ "hidden_chart_oracle_utility": 2.0,
1781
+ "outcome_ptr": 0.0,
1782
+ "predicted_margin": -0.006652955183872211,
1783
+ "proposal_oracle_success": 0.0,
1784
+ "proposal_oracle_utility": 0.22727864980697632,
1785
+ "selected_candidate_index": 5.0,
1786
+ "selected_success": 0.0,
1787
+ "selected_success_gain_over_base": 0.0,
1788
+ "selected_utility": 0.6000905632972717,
1789
+ "selected_utility_gain_over_base": 0.0,
1790
+ "selector_regret": 0.0,
1791
+ "success_selector_gap": 0.0,
1792
+ "success_support_gap": 1.0,
1793
+ "support_gap": 1.7727213501930237,
1794
+ "tau": 0.08716187110665122
1795
+ },
1796
+ "412102563/train0": {
1797
+ "base_success": 1.0,
1798
+ "base_utility": 2.0,
1799
+ "coverage": 0.0,
1800
+ "execute_generated": 0.0,
1801
+ "fallback_rate": 1.0,
1802
+ "hidden_chart_oracle_success": 1.0,
1803
+ "hidden_chart_oracle_utility": 2.0,
1804
+ "outcome_ptr": 0.0,
1805
+ "predicted_margin": 0.058708422896759524,
1806
+ "proposal_oracle_success": 1.0,
1807
+ "proposal_oracle_utility": 2.0,
1808
+ "selected_candidate_index": 3.0,
1809
+ "selected_success": 1.0,
1810
+ "selected_success_gain_over_base": 0.0,
1811
+ "selected_utility": 2.0,
1812
+ "selected_utility_gain_over_base": 0.0,
1813
+ "selector_regret": 0.0,
1814
+ "success_selector_gap": 0.0,
1815
+ "success_support_gap": 0.0,
1816
+ "support_gap": 0.0,
1817
+ "tau": 0.08716187110665122
1818
+ },
1819
+ "412102563/train1": {
1820
+ "base_success": 1.0,
1821
+ "base_utility": 2.0,
1822
+ "coverage": 0.0,
1823
+ "execute_generated": 0.0,
1824
+ "fallback_rate": 1.0,
1825
+ "hidden_chart_oracle_success": 1.0,
1826
+ "hidden_chart_oracle_utility": 2.0,
1827
+ "outcome_ptr": 0.0,
1828
+ "predicted_margin": 0.07614491367847959,
1829
+ "proposal_oracle_success": 1.0,
1830
+ "proposal_oracle_utility": 2.0,
1831
+ "selected_candidate_index": 7.0,
1832
+ "selected_success": 1.0,
1833
+ "selected_success_gain_over_base": 0.0,
1834
+ "selected_utility": 2.0,
1835
+ "selected_utility_gain_over_base": 0.0,
1836
+ "selector_regret": 0.0,
1837
+ "success_selector_gap": 0.0,
1838
+ "success_support_gap": 0.0,
1839
+ "support_gap": 0.0,
1840
+ "tau": 0.08716187110665122
1841
+ },
1842
+ "412102563/train2": {
1843
+ "base_success": 1.0,
1844
+ "base_utility": 2.0,
1845
+ "coverage": 1.0,
1846
+ "execute_generated": 1.0,
1847
+ "fallback_rate": 0.0,
1848
+ "hidden_chart_oracle_success": 1.0,
1849
+ "hidden_chart_oracle_utility": 2.0,
1850
+ "outcome_ptr": 0.0,
1851
+ "predicted_margin": 0.11173390380414704,
1852
+ "proposal_oracle_success": 1.0,
1853
+ "proposal_oracle_utility": 2.0,
1854
+ "selected_candidate_index": 4.0,
1855
+ "selected_success": 0.0,
1856
+ "selected_success_gain_over_base": -1.0,
1857
+ "selected_utility": 0.14267796277999878,
1858
+ "selected_utility_gain_over_base": -1.8573220372200012,
1859
+ "selector_regret": 1.8573220372200012,
1860
+ "success_selector_gap": 1.0,
1861
+ "success_support_gap": 0.0,
1862
+ "support_gap": 0.0,
1863
+ "tau": 0.08716187110665122
1864
+ },
1865
+ "440577711/train0": {
1866
+ "base_success": 0.0,
1867
+ "base_utility": 0.036152780055999756,
1868
+ "coverage": 1.0,
1869
+ "execute_generated": 1.0,
1870
+ "fallback_rate": 0.0,
1871
+ "hidden_chart_oracle_success": 1.0,
1872
+ "hidden_chart_oracle_utility": 2.0,
1873
+ "outcome_ptr": 1.0,
1874
+ "predicted_margin": 0.25426567309182174,
1875
+ "proposal_oracle_success": 1.0,
1876
+ "proposal_oracle_utility": 2.0,
1877
+ "selected_candidate_index": 1.0,
1878
+ "selected_success": 1.0,
1879
+ "selected_success_gain_over_base": 1.0,
1880
+ "selected_utility": 2.0,
1881
+ "selected_utility_gain_over_base": 1.9638472199440002,
1882
+ "selector_regret": 0.0,
1883
+ "success_selector_gap": 0.0,
1884
+ "success_support_gap": 0.0,
1885
+ "support_gap": 0.0,
1886
+ "tau": 0.08716187110665122
1887
+ },
1888
+ "440577711/train1": {
1889
+ "base_success": 0.0,
1890
+ "base_utility": 0.036152780055999756,
1891
+ "coverage": 1.0,
1892
+ "execute_generated": 1.0,
1893
+ "fallback_rate": 0.0,
1894
+ "hidden_chart_oracle_success": 1.0,
1895
+ "hidden_chart_oracle_utility": 2.0,
1896
+ "outcome_ptr": 1.0,
1897
+ "predicted_margin": 0.4007284384303781,
1898
+ "proposal_oracle_success": 1.0,
1899
+ "proposal_oracle_utility": 2.0,
1900
+ "selected_candidate_index": 7.0,
1901
+ "selected_success": 1.0,
1902
+ "selected_success_gain_over_base": 1.0,
1903
+ "selected_utility": 2.0,
1904
+ "selected_utility_gain_over_base": 1.9638472199440002,
1905
+ "selector_regret": 0.0,
1906
+ "success_selector_gap": 0.0,
1907
+ "success_support_gap": 0.0,
1908
+ "support_gap": 0.0,
1909
+ "tau": 0.08716187110665122
1910
+ },
1911
+ "440577711/train2": {
1912
+ "base_success": 0.0,
1913
+ "base_utility": 0.036152780055999756,
1914
+ "coverage": 1.0,
1915
+ "execute_generated": 1.0,
1916
+ "fallback_rate": 0.0,
1917
+ "hidden_chart_oracle_success": 1.0,
1918
+ "hidden_chart_oracle_utility": 2.0,
1919
+ "outcome_ptr": 1.0,
1920
+ "predicted_margin": 0.2441943745548885,
1921
+ "proposal_oracle_success": 1.0,
1922
+ "proposal_oracle_utility": 2.0,
1923
+ "selected_candidate_index": 1.0,
1924
+ "selected_success": 1.0,
1925
+ "selected_success_gain_over_base": 1.0,
1926
+ "selected_utility": 2.0,
1927
+ "selected_utility_gain_over_base": 1.9638472199440002,
1928
+ "selector_regret": 0.0,
1929
+ "success_selector_gap": 0.0,
1930
+ "success_support_gap": 0.0,
1931
+ "support_gap": 0.0,
1932
+ "tau": 0.08716187110665122
1933
+ },
1934
+ "441046872/train0": {
1935
+ "base_success": 0.0,
1936
+ "base_utility": 0.6999551653862,
1937
+ "coverage": 0.0,
1938
+ "execute_generated": 0.0,
1939
+ "fallback_rate": 1.0,
1940
+ "hidden_chart_oracle_success": 1.0,
1941
+ "hidden_chart_oracle_utility": 2.0,
1942
+ "outcome_ptr": 0.0,
1943
+ "predicted_margin": 0.039114097271145426,
1944
+ "proposal_oracle_success": 0.0,
1945
+ "proposal_oracle_utility": 0.5148025155067444,
1946
+ "selected_candidate_index": 0.0,
1947
+ "selected_success": 0.0,
1948
+ "selected_success_gain_over_base": 0.0,
1949
+ "selected_utility": 0.6999551653862,
1950
+ "selected_utility_gain_over_base": 0.0,
1951
+ "selector_regret": 0.0,
1952
+ "success_selector_gap": 0.0,
1953
+ "success_support_gap": 1.0,
1954
+ "support_gap": 1.4851974844932556,
1955
+ "tau": 0.08716187110665122
1956
+ },
1957
+ "441046872/train1": {
1958
+ "base_success": 0.0,
1959
+ "base_utility": 0.6999551653862,
1960
+ "coverage": 0.0,
1961
+ "execute_generated": 0.0,
1962
+ "fallback_rate": 1.0,
1963
+ "hidden_chart_oracle_success": 1.0,
1964
+ "hidden_chart_oracle_utility": 2.0,
1965
+ "outcome_ptr": 1.0,
1966
+ "predicted_margin": -0.043013989308341063,
1967
+ "proposal_oracle_success": 0.0,
1968
+ "proposal_oracle_utility": 0.7233588695526123,
1969
+ "selected_candidate_index": 2.0,
1970
+ "selected_success": 0.0,
1971
+ "selected_success_gain_over_base": 0.0,
1972
+ "selected_utility": 0.6999551653862,
1973
+ "selected_utility_gain_over_base": 0.0,
1974
+ "selector_regret": 0.023403704166412354,
1975
+ "success_selector_gap": 0.0,
1976
+ "success_support_gap": 1.0,
1977
+ "support_gap": 1.2766411304473877,
1978
+ "tau": 0.08716187110665122
1979
+ },
1980
+ "441046872/train2": {
1981
+ "base_success": 0.0,
1982
+ "base_utility": 0.6999551653862,
1983
+ "coverage": 0.0,
1984
+ "execute_generated": 0.0,
1985
+ "fallback_rate": 1.0,
1986
+ "hidden_chart_oracle_success": 1.0,
1987
+ "hidden_chart_oracle_utility": 2.0,
1988
+ "outcome_ptr": 1.0,
1989
+ "predicted_margin": 0.019269895539017243,
1990
+ "proposal_oracle_success": 0.0,
1991
+ "proposal_oracle_utility": 0.7130289077758789,
1992
+ "selected_candidate_index": 1.0,
1993
+ "selected_success": 0.0,
1994
+ "selected_success_gain_over_base": 0.0,
1995
+ "selected_utility": 0.6999551653862,
1996
+ "selected_utility_gain_over_base": 0.0,
1997
+ "selector_regret": 0.013073742389678955,
1998
+ "success_selector_gap": 0.0,
1999
+ "success_support_gap": 1.0,
2000
+ "support_gap": 1.286971092224121,
2001
+ "tau": 0.08716187110665122
2002
+ },
2003
+ "509430392/train0": {
2004
+ "base_success": 0.0,
2005
+ "base_utility": 2.2351741790771484e-07,
2006
+ "coverage": 1.0,
2007
+ "execute_generated": 1.0,
2008
+ "fallback_rate": 0.0,
2009
+ "hidden_chart_oracle_success": 1.0,
2010
+ "hidden_chart_oracle_utility": 2.0,
2011
+ "outcome_ptr": 1.0,
2012
+ "predicted_margin": 0.15209589793778405,
2013
+ "proposal_oracle_success": 1.0,
2014
+ "proposal_oracle_utility": 2.0,
2015
+ "selected_candidate_index": 4.0,
2016
+ "selected_success": 0.0,
2017
+ "selected_success_gain_over_base": 0.0,
2018
+ "selected_utility": 6.109476089477539e-07,
2019
+ "selected_utility_gain_over_base": 3.8743019104003906e-07,
2020
+ "selector_regret": 1.999999389052391,
2021
+ "success_selector_gap": 1.0,
2022
+ "success_support_gap": 0.0,
2023
+ "support_gap": 0.0,
2024
+ "tau": 0.08716187110665122
2025
+ },
2026
+ "509430392/train1": {
2027
+ "base_success": 0.0,
2028
+ "base_utility": 2.2351741790771484e-07,
2029
+ "coverage": 1.0,
2030
+ "execute_generated": 1.0,
2031
+ "fallback_rate": 0.0,
2032
+ "hidden_chart_oracle_success": 1.0,
2033
+ "hidden_chart_oracle_utility": 2.0,
2034
+ "outcome_ptr": 1.0,
2035
+ "predicted_margin": 0.20205288945859612,
2036
+ "proposal_oracle_success": 0.0,
2037
+ "proposal_oracle_utility": 6.164610385894775e-05,
2038
+ "selected_candidate_index": 2.0,
2039
+ "selected_success": 0.0,
2040
+ "selected_success_gain_over_base": 0.0,
2041
+ "selected_utility": 1.519918441772461e-06,
2042
+ "selected_utility_gain_over_base": 1.296401023864746e-06,
2043
+ "selector_regret": 6.012618541717529e-05,
2044
+ "success_selector_gap": 0.0,
2045
+ "success_support_gap": 1.0,
2046
+ "support_gap": 1.999938353896141,
2047
+ "tau": 0.08716187110665122
2048
+ },
2049
+ "509430392/train2": {
2050
+ "base_success": 0.0,
2051
+ "base_utility": 2.2351741790771484e-07,
2052
+ "coverage": 0.0,
2053
+ "execute_generated": 0.0,
2054
+ "fallback_rate": 1.0,
2055
+ "hidden_chart_oracle_success": 1.0,
2056
+ "hidden_chart_oracle_utility": 2.0,
2057
+ "outcome_ptr": 1.0,
2058
+ "predicted_margin": 0.068865962955827,
2059
+ "proposal_oracle_success": 0.0,
2060
+ "proposal_oracle_utility": 5.424022674560547e-06,
2061
+ "selected_candidate_index": 5.0,
2062
+ "selected_success": 0.0,
2063
+ "selected_success_gain_over_base": 0.0,
2064
+ "selected_utility": 2.2351741790771484e-07,
2065
+ "selected_utility_gain_over_base": 0.0,
2066
+ "selector_regret": 5.200505256652832e-06,
2067
+ "success_selector_gap": 0.0,
2068
+ "success_support_gap": 1.0,
2069
+ "support_gap": 1.9999945759773254,
2070
+ "tau": 0.08716187110665122
2071
+ },
2072
+ "585904130/train0": {
2073
+ "base_success": 0.0,
2074
+ "base_utility": 0.18171554803848267,
2075
+ "coverage": 0.0,
2076
+ "execute_generated": 0.0,
2077
+ "fallback_rate": 1.0,
2078
+ "hidden_chart_oracle_success": 1.0,
2079
+ "hidden_chart_oracle_utility": 2.0,
2080
+ "outcome_ptr": 1.0,
2081
+ "predicted_margin": -0.1377085493162327,
2082
+ "proposal_oracle_success": 1.0,
2083
+ "proposal_oracle_utility": 2.0,
2084
+ "selected_candidate_index": 0.0,
2085
+ "selected_success": 0.0,
2086
+ "selected_success_gain_over_base": 0.0,
2087
+ "selected_utility": 0.18171554803848267,
2088
+ "selected_utility_gain_over_base": 0.0,
2089
+ "selector_regret": 1.8182844519615173,
2090
+ "success_selector_gap": 1.0,
2091
+ "success_support_gap": 0.0,
2092
+ "support_gap": 0.0,
2093
+ "tau": 0.08716187110665122
2094
+ },
2095
+ "585904130/train1": {
2096
+ "base_success": 0.0,
2097
+ "base_utility": 0.18171554803848267,
2098
+ "coverage": 0.0,
2099
+ "execute_generated": 0.0,
2100
+ "fallback_rate": 1.0,
2101
+ "hidden_chart_oracle_success": 1.0,
2102
+ "hidden_chart_oracle_utility": 2.0,
2103
+ "outcome_ptr": 1.0,
2104
+ "predicted_margin": -0.1018982922151131,
2105
+ "proposal_oracle_success": 1.0,
2106
+ "proposal_oracle_utility": 2.0,
2107
+ "selected_candidate_index": 0.0,
2108
+ "selected_success": 0.0,
2109
+ "selected_success_gain_over_base": 0.0,
2110
+ "selected_utility": 0.18171554803848267,
2111
+ "selected_utility_gain_over_base": 0.0,
2112
+ "selector_regret": 1.8182844519615173,
2113
+ "success_selector_gap": 1.0,
2114
+ "success_support_gap": 0.0,
2115
+ "support_gap": 0.0,
2116
+ "tau": 0.08716187110665122
2117
+ },
2118
+ "585904130/train2": {
2119
+ "base_success": 0.0,
2120
+ "base_utility": 0.18171554803848267,
2121
+ "coverage": 0.0,
2122
+ "execute_generated": 0.0,
2123
+ "fallback_rate": 1.0,
2124
+ "hidden_chart_oracle_success": 1.0,
2125
+ "hidden_chart_oracle_utility": 2.0,
2126
+ "outcome_ptr": 1.0,
2127
+ "predicted_margin": -0.11283069280179209,
2128
+ "proposal_oracle_success": 1.0,
2129
+ "proposal_oracle_utility": 2.0,
2130
+ "selected_candidate_index": 0.0,
2131
+ "selected_success": 0.0,
2132
+ "selected_success_gain_over_base": 0.0,
2133
+ "selected_utility": 0.18171554803848267,
2134
+ "selected_utility_gain_over_base": 0.0,
2135
+ "selector_regret": 1.8182844519615173,
2136
+ "success_selector_gap": 1.0,
2137
+ "success_support_gap": 0.0,
2138
+ "support_gap": 0.0,
2139
+ "tau": 0.08716187110665122
2140
+ },
2141
+ "608016290/train0": {
2142
+ "base_success": 0.0,
2143
+ "base_utility": 0.725701630115509,
2144
+ "coverage": 0.0,
2145
+ "execute_generated": 0.0,
2146
+ "fallback_rate": 1.0,
2147
+ "hidden_chart_oracle_success": 1.0,
2148
+ "hidden_chart_oracle_utility": 2.0,
2149
+ "outcome_ptr": 1.0,
2150
+ "predicted_margin": 0.0411035483314815,
2151
+ "proposal_oracle_success": 1.0,
2152
+ "proposal_oracle_utility": 2.0,
2153
+ "selected_candidate_index": 3.0,
2154
+ "selected_success": 0.0,
2155
+ "selected_success_gain_over_base": 0.0,
2156
+ "selected_utility": 0.725701630115509,
2157
+ "selected_utility_gain_over_base": 0.0,
2158
+ "selector_regret": 1.274298369884491,
2159
+ "success_selector_gap": 1.0,
2160
+ "success_support_gap": 0.0,
2161
+ "support_gap": 0.0,
2162
+ "tau": 0.08716187110665122
2163
+ },
2164
+ "608016290/train1": {
2165
+ "base_success": 0.0,
2166
+ "base_utility": 0.725701630115509,
2167
+ "coverage": 0.0,
2168
+ "execute_generated": 0.0,
2169
+ "fallback_rate": 1.0,
2170
+ "hidden_chart_oracle_success": 1.0,
2171
+ "hidden_chart_oracle_utility": 2.0,
2172
+ "outcome_ptr": 1.0,
2173
+ "predicted_margin": 0.022783867776710112,
2174
+ "proposal_oracle_success": 1.0,
2175
+ "proposal_oracle_utility": 2.0,
2176
+ "selected_candidate_index": 1.0,
2177
+ "selected_success": 0.0,
2178
+ "selected_success_gain_over_base": 0.0,
2179
+ "selected_utility": 0.725701630115509,
2180
+ "selected_utility_gain_over_base": 0.0,
2181
+ "selector_regret": 1.274298369884491,
2182
+ "success_selector_gap": 1.0,
2183
+ "success_support_gap": 0.0,
2184
+ "support_gap": 0.0,
2185
+ "tau": 0.08716187110665122
2186
+ },
2187
+ "608016290/train2": {
2188
+ "base_success": 0.0,
2189
+ "base_utility": 0.725701630115509,
2190
+ "coverage": 0.0,
2191
+ "execute_generated": 0.0,
2192
+ "fallback_rate": 1.0,
2193
+ "hidden_chart_oracle_success": 1.0,
2194
+ "hidden_chart_oracle_utility": 2.0,
2195
+ "outcome_ptr": 0.0,
2196
+ "predicted_margin": 0.0345396940539498,
2197
+ "proposal_oracle_success": 0.0,
2198
+ "proposal_oracle_utility": 0.6825780272483826,
2199
+ "selected_candidate_index": 2.0,
2200
+ "selected_success": 0.0,
2201
+ "selected_success_gain_over_base": 0.0,
2202
+ "selected_utility": 0.725701630115509,
2203
+ "selected_utility_gain_over_base": 0.0,
2204
+ "selector_regret": 0.0,
2205
+ "success_selector_gap": 0.0,
2206
+ "success_support_gap": 1.0,
2207
+ "support_gap": 1.3174219727516174,
2208
+ "tau": 0.08716187110665122
2209
+ },
2210
+ "626486813/train0": {
2211
+ "base_success": 0.0,
2212
+ "base_utility": 0.00509834272088483,
2213
+ "coverage": 0.5,
2214
+ "execute_generated": 0.5,
2215
+ "fallback_rate": 0.5,
2216
+ "hidden_chart_oracle_success": 0.0,
2217
+ "hidden_chart_oracle_utility": 0.16558238118886948,
2218
+ "outcome_ptr": 1.0,
2219
+ "predicted_margin": 0.08596848299541204,
2220
+ "proposal_oracle_success": 0.0,
2221
+ "proposal_oracle_utility": 0.043898362666368484,
2222
+ "selected_candidate_index": 5.0,
2223
+ "selected_success": 0.0,
2224
+ "selected_success_gain_over_base": 0.0,
2225
+ "selected_utility": 0.007657766109332442,
2226
+ "selected_utility_gain_over_base": 0.0025594233884476125,
2227
+ "selector_regret": 0.03624059655703604,
2228
+ "success_selector_gap": 0.0,
2229
+ "success_support_gap": 0.0,
2230
+ "support_gap": 0.12241578102111816,
2231
+ "tau": 0.08716187110665122
2232
+ },
2233
+ "626486813/train1": {
2234
+ "base_success": 0.0,
2235
+ "base_utility": 0.00509834272088483,
2236
+ "coverage": 1.0,
2237
+ "execute_generated": 1.0,
2238
+ "fallback_rate": 0.0,
2239
+ "hidden_chart_oracle_success": 0.0,
2240
+ "hidden_chart_oracle_utility": 0.16558238118886948,
2241
+ "outcome_ptr": 1.0,
2242
+ "predicted_margin": 0.1404318861185406,
2243
+ "proposal_oracle_success": 0.0,
2244
+ "proposal_oracle_utility": 0.030555277597159147,
2245
+ "selected_candidate_index": 4.0,
2246
+ "selected_success": 0.0,
2247
+ "selected_success_gain_over_base": 0.0,
2248
+ "selected_utility": 0.019377360120415688,
2249
+ "selected_utility_gain_over_base": 0.014279017399530858,
2250
+ "selector_regret": 0.01117791747674346,
2251
+ "success_selector_gap": 0.0,
2252
+ "success_support_gap": 0.0,
2253
+ "support_gap": 0.13502710359171033,
2254
+ "tau": 0.08716187110665122
2255
+ },
2256
+ "626486813/train2": {
2257
+ "base_success": 0.0,
2258
+ "base_utility": 0.00509834272088483,
2259
+ "coverage": 0.0,
2260
+ "execute_generated": 0.0,
2261
+ "fallback_rate": 1.0,
2262
+ "hidden_chart_oracle_success": 0.0,
2263
+ "hidden_chart_oracle_utility": 0.16558238118886948,
2264
+ "outcome_ptr": 1.0,
2265
+ "predicted_margin": 0.02041522092884359,
2266
+ "proposal_oracle_success": 0.0,
2267
+ "proposal_oracle_utility": 0.03856666944921017,
2268
+ "selected_candidate_index": 1.0,
2269
+ "selected_success": 0.0,
2270
+ "selected_success_gain_over_base": 0.0,
2271
+ "selected_utility": 0.00509834272088483,
2272
+ "selected_utility_gain_over_base": 0.0,
2273
+ "selector_regret": 0.03346832672832534,
2274
+ "success_selector_gap": 0.0,
2275
+ "success_support_gap": 0.0,
2276
+ "support_gap": 0.1270157117396593,
2277
+ "tau": 0.08716187110665122
2278
+ },
2279
+ "636680137/train0": {
2280
+ "base_success": 0.0,
2281
+ "base_utility": 0.08313504606485367,
2282
+ "coverage": 1.0,
2283
+ "execute_generated": 1.0,
2284
+ "fallback_rate": 0.0,
2285
+ "hidden_chart_oracle_success": 0.0,
2286
+ "hidden_chart_oracle_utility": 0.17056432366371155,
2287
+ "outcome_ptr": 1.0,
2288
+ "predicted_margin": 0.1210330001510371,
2289
+ "proposal_oracle_success": 0.0,
2290
+ "proposal_oracle_utility": 0.08826130628585815,
2291
+ "selected_candidate_index": 1.0,
2292
+ "selected_success": 0.0,
2293
+ "selected_success_gain_over_base": 0.0,
2294
+ "selected_utility": 0.08826130628585815,
2295
+ "selected_utility_gain_over_base": 0.005126260221004486,
2296
+ "selector_regret": 0.0,
2297
+ "success_selector_gap": 0.0,
2298
+ "success_support_gap": 0.0,
2299
+ "support_gap": 0.0823030173778534,
2300
+ "tau": 0.08716187110665122
2301
+ },
2302
+ "636680137/train1": {
2303
+ "base_success": 0.0,
2304
+ "base_utility": 0.08313504606485367,
2305
+ "coverage": 0.0,
2306
+ "execute_generated": 0.0,
2307
+ "fallback_rate": 1.0,
2308
+ "hidden_chart_oracle_success": 0.0,
2309
+ "hidden_chart_oracle_utility": 0.17056432366371155,
2310
+ "outcome_ptr": 0.0,
2311
+ "predicted_margin": 0.056941927161848475,
2312
+ "proposal_oracle_success": 0.0,
2313
+ "proposal_oracle_utility": 0.06741118431091309,
2314
+ "selected_candidate_index": 1.0,
2315
+ "selected_success": 0.0,
2316
+ "selected_success_gain_over_base": 0.0,
2317
+ "selected_utility": 0.08313504606485367,
2318
+ "selected_utility_gain_over_base": 0.0,
2319
+ "selector_regret": 0.0,
2320
+ "success_selector_gap": 0.0,
2321
+ "success_support_gap": 0.0,
2322
+ "support_gap": 0.10315313935279846,
2323
+ "tau": 0.08716187110665122
2324
+ },
2325
+ "636680137/train2": {
2326
+ "base_success": 0.0,
2327
+ "base_utility": 0.08313504606485367,
2328
+ "coverage": 0.0,
2329
+ "execute_generated": 0.0,
2330
+ "fallback_rate": 1.0,
2331
+ "hidden_chart_oracle_success": 0.0,
2332
+ "hidden_chart_oracle_utility": 0.17056432366371155,
2333
+ "outcome_ptr": 0.0,
2334
+ "predicted_margin": 0.06045216440073517,
2335
+ "proposal_oracle_success": 0.0,
2336
+ "proposal_oracle_utility": 0.0756516233086586,
2337
+ "selected_candidate_index": 7.0,
2338
+ "selected_success": 0.0,
2339
+ "selected_success_gain_over_base": 0.0,
2340
+ "selected_utility": 0.08313504606485367,
2341
+ "selected_utility_gain_over_base": 0.0,
2342
+ "selector_regret": 0.0,
2343
+ "success_selector_gap": 0.0,
2344
+ "success_support_gap": 0.0,
2345
+ "support_gap": 0.09491270035505295,
2346
+ "tau": 0.08716187110665122
2347
+ },
2348
+ "655884605/train0": {
2349
+ "base_success": 0.0,
2350
+ "base_utility": 0.16300667822360992,
2351
+ "coverage": 0.0,
2352
+ "execute_generated": 0.0,
2353
+ "fallback_rate": 1.0,
2354
+ "hidden_chart_oracle_success": 1.0,
2355
+ "hidden_chart_oracle_utility": 2.0,
2356
+ "outcome_ptr": 0.0,
2357
+ "predicted_margin": 0.0015108551590233277,
2358
+ "proposal_oracle_success": 0.0,
2359
+ "proposal_oracle_utility": 0.13732868432998657,
2360
+ "selected_candidate_index": 6.0,
2361
+ "selected_success": 0.0,
2362
+ "selected_success_gain_over_base": 0.0,
2363
+ "selected_utility": 0.16300667822360992,
2364
+ "selected_utility_gain_over_base": 0.0,
2365
+ "selector_regret": 0.0,
2366
+ "success_selector_gap": 0.0,
2367
+ "success_support_gap": 1.0,
2368
+ "support_gap": 1.8626713156700134,
2369
+ "tau": 0.08716187110665122
2370
+ },
2371
+ "655884605/train1": {
2372
+ "base_success": 0.0,
2373
+ "base_utility": 0.16300667822360992,
2374
+ "coverage": 0.0,
2375
+ "execute_generated": 0.0,
2376
+ "fallback_rate": 1.0,
2377
+ "hidden_chart_oracle_success": 1.0,
2378
+ "hidden_chart_oracle_utility": 2.0,
2379
+ "outcome_ptr": 1.0,
2380
+ "predicted_margin": -0.003742232991734387,
2381
+ "proposal_oracle_success": 0.0,
2382
+ "proposal_oracle_utility": 0.1651531159877777,
2383
+ "selected_candidate_index": 7.0,
2384
+ "selected_success": 0.0,
2385
+ "selected_success_gain_over_base": 0.0,
2386
+ "selected_utility": 0.16300667822360992,
2387
+ "selected_utility_gain_over_base": 0.0,
2388
+ "selector_regret": 0.0021464377641677856,
2389
+ "success_selector_gap": 0.0,
2390
+ "success_support_gap": 1.0,
2391
+ "support_gap": 1.8348468840122223,
2392
+ "tau": 0.08716187110665122
2393
+ },
2394
+ "655884605/train2": {
2395
+ "base_success": 0.0,
2396
+ "base_utility": 0.16300667822360992,
2397
+ "coverage": 0.0,
2398
+ "execute_generated": 0.0,
2399
+ "fallback_rate": 1.0,
2400
+ "hidden_chart_oracle_success": 1.0,
2401
+ "hidden_chart_oracle_utility": 2.0,
2402
+ "outcome_ptr": 1.0,
2403
+ "predicted_margin": 0.05371800066830732,
2404
+ "proposal_oracle_success": 0.0,
2405
+ "proposal_oracle_utility": 0.21447330713272095,
2406
+ "selected_candidate_index": 7.0,
2407
+ "selected_success": 0.0,
2408
+ "selected_success_gain_over_base": 0.0,
2409
+ "selected_utility": 0.16300667822360992,
2410
+ "selected_utility_gain_over_base": 0.0,
2411
+ "selector_regret": 0.05146662890911102,
2412
+ "success_selector_gap": 0.0,
2413
+ "success_support_gap": 1.0,
2414
+ "support_gap": 1.785526692867279,
2415
+ "tau": 0.08716187110665122
2416
+ },
2417
+ "672328262/train0": {
2418
+ "base_success": 1.0,
2419
+ "base_utility": 2.0,
2420
+ "coverage": 0.0,
2421
+ "execute_generated": 0.0,
2422
+ "fallback_rate": 1.0,
2423
+ "hidden_chart_oracle_success": 1.0,
2424
+ "hidden_chart_oracle_utility": 2.0,
2425
+ "outcome_ptr": 0.0,
2426
+ "predicted_margin": -0.03328084209105449,
2427
+ "proposal_oracle_success": 1.0,
2428
+ "proposal_oracle_utility": 2.0,
2429
+ "selected_candidate_index": 1.0,
2430
+ "selected_success": 1.0,
2431
+ "selected_success_gain_over_base": 0.0,
2432
+ "selected_utility": 2.0,
2433
+ "selected_utility_gain_over_base": 0.0,
2434
+ "selector_regret": 0.0,
2435
+ "success_selector_gap": 0.0,
2436
+ "success_support_gap": 0.0,
2437
+ "support_gap": 0.0,
2438
+ "tau": 0.08716187110665122
2439
+ },
2440
+ "672328262/train1": {
2441
+ "base_success": 1.0,
2442
+ "base_utility": 2.0,
2443
+ "coverage": 0.0,
2444
+ "execute_generated": 0.0,
2445
+ "fallback_rate": 1.0,
2446
+ "hidden_chart_oracle_success": 1.0,
2447
+ "hidden_chart_oracle_utility": 2.0,
2448
+ "outcome_ptr": 0.0,
2449
+ "predicted_margin": -0.003607790388386746,
2450
+ "proposal_oracle_success": 1.0,
2451
+ "proposal_oracle_utility": 2.0,
2452
+ "selected_candidate_index": 0.0,
2453
+ "selected_success": 1.0,
2454
+ "selected_success_gain_over_base": 0.0,
2455
+ "selected_utility": 2.0,
2456
+ "selected_utility_gain_over_base": 0.0,
2457
+ "selector_regret": 0.0,
2458
+ "success_selector_gap": 0.0,
2459
+ "success_support_gap": 0.0,
2460
+ "support_gap": 0.0,
2461
+ "tau": 0.08716187110665122
2462
+ },
2463
+ "672328262/train2": {
2464
+ "base_success": 1.0,
2465
+ "base_utility": 2.0,
2466
+ "coverage": 0.0,
2467
+ "execute_generated": 0.0,
2468
+ "fallback_rate": 1.0,
2469
+ "hidden_chart_oracle_success": 1.0,
2470
+ "hidden_chart_oracle_utility": 2.0,
2471
+ "outcome_ptr": 0.0,
2472
+ "predicted_margin": -0.06253280565388909,
2473
+ "proposal_oracle_success": 1.0,
2474
+ "proposal_oracle_utility": 2.0,
2475
+ "selected_candidate_index": 1.0,
2476
+ "selected_success": 1.0,
2477
+ "selected_success_gain_over_base": 0.0,
2478
+ "selected_utility": 2.0,
2479
+ "selected_utility_gain_over_base": 0.0,
2480
+ "selector_regret": 0.0,
2481
+ "success_selector_gap": 0.0,
2482
+ "success_support_gap": 0.0,
2483
+ "support_gap": 0.0,
2484
+ "tau": 0.08716187110665122
2485
+ },
2486
+ "701667486/train0": {
2487
+ "base_success": 1.0,
2488
+ "base_utility": 2.0,
2489
+ "coverage": 0.0,
2490
+ "execute_generated": 0.0,
2491
+ "fallback_rate": 1.0,
2492
+ "hidden_chart_oracle_success": 1.0,
2493
+ "hidden_chart_oracle_utility": 2.0,
2494
+ "outcome_ptr": 0.0,
2495
+ "predicted_margin": 0.07238459945730863,
2496
+ "proposal_oracle_success": 1.0,
2497
+ "proposal_oracle_utility": 2.0,
2498
+ "selected_candidate_index": 2.0,
2499
+ "selected_success": 1.0,
2500
+ "selected_success_gain_over_base": 0.0,
2501
+ "selected_utility": 2.0,
2502
+ "selected_utility_gain_over_base": 0.0,
2503
+ "selector_regret": 0.0,
2504
+ "success_selector_gap": 0.0,
2505
+ "success_support_gap": 0.0,
2506
+ "support_gap": 0.0,
2507
+ "tau": 0.08716187110665122
2508
+ },
2509
+ "701667486/train1": {
2510
+ "base_success": 1.0,
2511
+ "base_utility": 2.0,
2512
+ "coverage": 1.0,
2513
+ "execute_generated": 1.0,
2514
+ "fallback_rate": 0.0,
2515
+ "hidden_chart_oracle_success": 1.0,
2516
+ "hidden_chart_oracle_utility": 2.0,
2517
+ "outcome_ptr": 0.0,
2518
+ "predicted_margin": 0.1020198858265147,
2519
+ "proposal_oracle_success": 1.0,
2520
+ "proposal_oracle_utility": 2.0,
2521
+ "selected_candidate_index": 4.0,
2522
+ "selected_success": 1.0,
2523
+ "selected_success_gain_over_base": 0.0,
2524
+ "selected_utility": 2.0,
2525
+ "selected_utility_gain_over_base": 0.0,
2526
+ "selector_regret": 0.0,
2527
+ "success_selector_gap": 0.0,
2528
+ "success_support_gap": 0.0,
2529
+ "support_gap": 0.0,
2530
+ "tau": 0.08716187110665122
2531
+ },
2532
+ "701667486/train2": {
2533
+ "base_success": 1.0,
2534
+ "base_utility": 2.0,
2535
+ "coverage": 1.0,
2536
+ "execute_generated": 1.0,
2537
+ "fallback_rate": 0.0,
2538
+ "hidden_chart_oracle_success": 1.0,
2539
+ "hidden_chart_oracle_utility": 2.0,
2540
+ "outcome_ptr": 0.0,
2541
+ "predicted_margin": 0.15804603348351798,
2542
+ "proposal_oracle_success": 1.0,
2543
+ "proposal_oracle_utility": 2.0,
2544
+ "selected_candidate_index": 3.0,
2545
+ "selected_success": 1.0,
2546
+ "selected_success_gain_over_base": 0.0,
2547
+ "selected_utility": 2.0,
2548
+ "selected_utility_gain_over_base": 0.0,
2549
+ "selector_regret": 0.0,
2550
+ "success_selector_gap": 0.0,
2551
+ "success_support_gap": 0.0,
2552
+ "support_gap": 0.0,
2553
+ "tau": 0.08716187110665122
2554
+ },
2555
+ "747260437/train0": {
2556
+ "base_success": 1.0,
2557
+ "base_utility": 2.0,
2558
+ "coverage": 0.0,
2559
+ "execute_generated": 0.0,
2560
+ "fallback_rate": 1.0,
2561
+ "hidden_chart_oracle_success": 0.0,
2562
+ "hidden_chart_oracle_utility": 0.2786632478237152,
2563
+ "outcome_ptr": 0.0,
2564
+ "predicted_margin": -0.11634362569281631,
2565
+ "proposal_oracle_success": 0.0,
2566
+ "proposal_oracle_utility": 0.10839563608169556,
2567
+ "selected_candidate_index": 4.0,
2568
+ "selected_success": 1.0,
2569
+ "selected_success_gain_over_base": 0.0,
2570
+ "selected_utility": 2.0,
2571
+ "selected_utility_gain_over_base": 0.0,
2572
+ "selector_regret": 0.0,
2573
+ "success_selector_gap": 0.0,
2574
+ "success_support_gap": 0.0,
2575
+ "support_gap": 0.17026761174201965,
2576
+ "tau": 0.08716187110665122
2577
+ },
2578
+ "747260437/train1": {
2579
+ "base_success": 1.0,
2580
+ "base_utility": 2.0,
2581
+ "coverage": 0.0,
2582
+ "execute_generated": 0.0,
2583
+ "fallback_rate": 1.0,
2584
+ "hidden_chart_oracle_success": 0.0,
2585
+ "hidden_chart_oracle_utility": 0.2786632478237152,
2586
+ "outcome_ptr": 0.0,
2587
+ "predicted_margin": -0.04343225657198474,
2588
+ "proposal_oracle_success": 0.0,
2589
+ "proposal_oracle_utility": 0.05267423391342163,
2590
+ "selected_candidate_index": 3.0,
2591
+ "selected_success": 1.0,
2592
+ "selected_success_gain_over_base": 0.0,
2593
+ "selected_utility": 2.0,
2594
+ "selected_utility_gain_over_base": 0.0,
2595
+ "selector_regret": 0.0,
2596
+ "success_selector_gap": 0.0,
2597
+ "success_support_gap": 0.0,
2598
+ "support_gap": 0.22598901391029358,
2599
+ "tau": 0.08716187110665122
2600
+ },
2601
+ "747260437/train2": {
2602
+ "base_success": 1.0,
2603
+ "base_utility": 2.0,
2604
+ "coverage": 0.0,
2605
+ "execute_generated": 0.0,
2606
+ "fallback_rate": 1.0,
2607
+ "hidden_chart_oracle_success": 0.0,
2608
+ "hidden_chart_oracle_utility": 0.2786632478237152,
2609
+ "outcome_ptr": 0.0,
2610
+ "predicted_margin": -0.04436313100748629,
2611
+ "proposal_oracle_success": 0.0,
2612
+ "proposal_oracle_utility": 0.12510187923908234,
2613
+ "selected_candidate_index": 4.0,
2614
+ "selected_success": 1.0,
2615
+ "selected_success_gain_over_base": 0.0,
2616
+ "selected_utility": 2.0,
2617
+ "selected_utility_gain_over_base": 0.0,
2618
+ "selector_regret": 0.0,
2619
+ "success_selector_gap": 0.0,
2620
+ "success_support_gap": 0.0,
2621
+ "support_gap": 0.15356136858463287,
2622
+ "tau": 0.08716187110665122
2623
+ },
2624
+ "750448296/train0": {
2625
+ "base_success": 0.0,
2626
+ "base_utility": 0.11759314934412639,
2627
+ "coverage": 0.3333333333333333,
2628
+ "execute_generated": 0.3333333333333333,
2629
+ "fallback_rate": 0.6666666666666666,
2630
+ "hidden_chart_oracle_success": 0.6666666666666666,
2631
+ "hidden_chart_oracle_utility": 1.4079964955647786,
2632
+ "outcome_ptr": 0.6666666666666666,
2633
+ "predicted_margin": -0.1540811675254832,
2634
+ "proposal_oracle_success": 0.3333333333333333,
2635
+ "proposal_oracle_utility": 0.7526603279014429,
2636
+ "selected_candidate_index": 2.3333333333333335,
2637
+ "selected_success": 0.0,
2638
+ "selected_success_gain_over_base": 0.0,
2639
+ "selected_utility": 0.1003921168545882,
2640
+ "selected_utility_gain_over_base": -0.017201032489538193,
2641
+ "selector_regret": 0.6522682110468546,
2642
+ "success_selector_gap": 0.3333333333333333,
2643
+ "success_support_gap": 0.3333333333333333,
2644
+ "support_gap": 0.6553361676633358,
2645
+ "tau": 0.08716187110665123
2646
+ },
2647
+ "750448296/train1": {
2648
+ "base_success": 0.0,
2649
+ "base_utility": 0.11759314934412639,
2650
+ "coverage": 0.0,
2651
+ "execute_generated": 0.0,
2652
+ "fallback_rate": 1.0,
2653
+ "hidden_chart_oracle_success": 0.6666666666666666,
2654
+ "hidden_chart_oracle_utility": 1.4079964955647786,
2655
+ "outcome_ptr": 1.0,
2656
+ "predicted_margin": -0.19815602200806393,
2657
+ "proposal_oracle_success": 0.6666666666666666,
2658
+ "proposal_oracle_utility": 1.3705032741030057,
2659
+ "selected_candidate_index": 4.0,
2660
+ "selected_success": 0.0,
2661
+ "selected_success_gain_over_base": 0.0,
2662
+ "selected_utility": 0.11759314934412639,
2663
+ "selected_utility_gain_over_base": 0.0,
2664
+ "selector_regret": 1.2529101247588794,
2665
+ "success_selector_gap": 0.6666666666666666,
2666
+ "success_support_gap": 0.0,
2667
+ "support_gap": 0.03749322146177292,
2668
+ "tau": 0.08716187110665123
2669
+ },
2670
+ "750448296/train2": {
2671
+ "base_success": 0.0,
2672
+ "base_utility": 0.11759314934412639,
2673
+ "coverage": 0.3333333333333333,
2674
+ "execute_generated": 0.3333333333333333,
2675
+ "fallback_rate": 0.6666666666666666,
2676
+ "hidden_chart_oracle_success": 0.6666666666666666,
2677
+ "hidden_chart_oracle_utility": 1.4079964955647786,
2678
+ "outcome_ptr": 0.6666666666666666,
2679
+ "predicted_margin": -0.21116734030360362,
2680
+ "proposal_oracle_success": 0.6666666666666666,
2681
+ "proposal_oracle_utility": 1.349846323331197,
2682
+ "selected_candidate_index": 2.3333333333333335,
2683
+ "selected_success": 0.0,
2684
+ "selected_success_gain_over_base": 0.0,
2685
+ "selected_utility": 0.10310697555541992,
2686
+ "selected_utility_gain_over_base": -0.014486173788706461,
2687
+ "selector_regret": 1.246739347775777,
2688
+ "success_selector_gap": 0.6666666666666666,
2689
+ "success_support_gap": 0.0,
2690
+ "support_gap": 0.05815017223358154,
2691
+ "tau": 0.08716187110665123
2692
+ },
2693
+ "767747424/train0": {
2694
+ "base_success": 0.25,
2695
+ "base_utility": 0.5731993420049548,
2696
+ "coverage": 0.0,
2697
+ "execute_generated": 0.0,
2698
+ "fallback_rate": 1.0,
2699
+ "hidden_chart_oracle_success": 0.25,
2700
+ "hidden_chart_oracle_utility": 0.6737305596470833,
2701
+ "outcome_ptr": 0.5,
2702
+ "predicted_margin": -0.08627011556643746,
2703
+ "proposal_oracle_success": 0.5,
2704
+ "proposal_oracle_utility": 1.0424291878007352,
2705
+ "selected_candidate_index": 4.5,
2706
+ "selected_success": 0.25,
2707
+ "selected_success_gain_over_base": 0.0,
2708
+ "selected_utility": 0.5731993420049548,
2709
+ "selected_utility_gain_over_base": 0.0,
2710
+ "selector_regret": 0.4779601227492094,
2711
+ "success_selector_gap": 0.25,
2712
+ "success_support_gap": 0.0,
2713
+ "support_gap": 0.06562284985557199,
2714
+ "tau": 0.08716187110665122
2715
+ },
2716
+ "767747424/train1": {
2717
+ "base_success": 0.25,
2718
+ "base_utility": 0.5731993420049548,
2719
+ "coverage": 0.25,
2720
+ "execute_generated": 0.25,
2721
+ "fallback_rate": 0.75,
2722
+ "hidden_chart_oracle_success": 0.25,
2723
+ "hidden_chart_oracle_utility": 0.6737305596470833,
2724
+ "outcome_ptr": 0.25,
2725
+ "predicted_margin": -0.048375800756182,
2726
+ "proposal_oracle_success": 0.25,
2727
+ "proposal_oracle_utility": 0.5951075805351138,
2728
+ "selected_candidate_index": 2.75,
2729
+ "selected_success": 0.25,
2730
+ "selected_success_gain_over_base": 0.0,
2731
+ "selected_utility": 0.5928727509453893,
2732
+ "selected_utility_gain_over_base": 0.019673408940434456,
2733
+ "selector_regret": 0.01333707571029663,
2734
+ "success_selector_gap": 0.0,
2735
+ "success_support_gap": 0.25,
2736
+ "support_gap": 0.5129444571211934,
2737
+ "tau": 0.08716187110665122
2738
+ },
2739
+ "767747424/train2": {
2740
+ "base_success": 0.25,
2741
+ "base_utility": 0.5731993420049548,
2742
+ "coverage": 0.0,
2743
+ "execute_generated": 0.0,
2744
+ "fallback_rate": 1.0,
2745
+ "hidden_chart_oracle_success": 0.25,
2746
+ "hidden_chart_oracle_utility": 0.6737305596470833,
2747
+ "outcome_ptr": 0.5,
2748
+ "predicted_margin": -0.012019933895878576,
2749
+ "proposal_oracle_success": 0.25,
2750
+ "proposal_oracle_utility": 0.6134321661666036,
2751
+ "selected_candidate_index": 3.0,
2752
+ "selected_success": 0.25,
2753
+ "selected_success_gain_over_base": 0.0,
2754
+ "selected_utility": 0.5731993420049548,
2755
+ "selected_utility_gain_over_base": 0.0,
2756
+ "selector_regret": 0.04225692339241505,
2757
+ "success_selector_gap": 0.0,
2758
+ "success_support_gap": 0.25,
2759
+ "support_gap": 0.49461987148970366,
2760
+ "tau": 0.08716187110665122
2761
+ }
2762
+ }
workspace/runs/ctt_base_context_obs_learned_dominance_context_envclip_train_to_test/metrics_by_task.json ADDED
@@ -0,0 +1,117 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "LiftPegUpright-v1": {
3
+ "base_success": 0.2,
4
+ "base_utility": 0.9498705387115478,
5
+ "coverage": 0.06666666666666667,
6
+ "execute_generated": 0.06666666666666667,
7
+ "fallback_rate": 0.9333333333333333,
8
+ "hidden_chart_oracle_success": 1.0,
9
+ "hidden_chart_oracle_utility": 2.0,
10
+ "outcome_ptr": 0.4666666666666667,
11
+ "predicted_margin": 0.005682207003903464,
12
+ "proposal_oracle_success": 0.4666666666666667,
13
+ "proposal_oracle_utility": 1.2244940986235937,
14
+ "selected_candidate_index": 3.2666666666666666,
15
+ "selected_success": 0.2,
16
+ "selected_success_gain_over_base": 0.0,
17
+ "selected_utility": 0.9233068734407425,
18
+ "selected_utility_gain_over_base": -0.026563665270805357,
19
+ "selector_regret": 0.42827867666880287,
20
+ "success_selector_gap": 0.3333333333333333,
21
+ "success_support_gap": 0.5333333333333333,
22
+ "support_gap": 0.7755059013764064,
23
+ "tau": 0.08716187110665123
24
+ },
25
+ "PickCube-v1": {
26
+ "base_success": 0.0,
27
+ "base_utility": 0.5839591026306152,
28
+ "coverage": 0.3333333333333333,
29
+ "execute_generated": 0.3333333333333333,
30
+ "fallback_rate": 0.6666666666666666,
31
+ "hidden_chart_oracle_success": 1.0,
32
+ "hidden_chart_oracle_utility": 2.0,
33
+ "outcome_ptr": 0.0,
34
+ "predicted_margin": -0.01662357846984096,
35
+ "proposal_oracle_success": 0.0,
36
+ "proposal_oracle_utility": 0.13425676027933756,
37
+ "selected_candidate_index": 4.0,
38
+ "selected_success": 0.0,
39
+ "selected_success_gain_over_base": 0.0,
40
+ "selected_utility": 0.4258223846554756,
41
+ "selected_utility_gain_over_base": -0.15813671797513962,
42
+ "selector_regret": 0.013274706900119781,
43
+ "success_selector_gap": 0.0,
44
+ "success_support_gap": 1.0,
45
+ "support_gap": 1.8657432397206624,
46
+ "tau": 0.08716187110665123
47
+ },
48
+ "PullCube-v1": {
49
+ "base_success": 0.36,
50
+ "base_utility": 0.7751704612234608,
51
+ "coverage": 0.22666666666666666,
52
+ "execute_generated": 0.22666666666666666,
53
+ "fallback_rate": 0.7733333333333333,
54
+ "hidden_chart_oracle_success": 0.52,
55
+ "hidden_chart_oracle_utility": 1.1467878860235214,
56
+ "outcome_ptr": 0.44,
57
+ "predicted_margin": -0.026633474795958578,
58
+ "proposal_oracle_success": 0.4666666666666667,
59
+ "proposal_oracle_utility": 0.9816135169876119,
60
+ "selected_candidate_index": 3.0,
61
+ "selected_success": 0.36,
62
+ "selected_success_gain_over_base": 0.0,
63
+ "selected_utility": 0.7728006126933421,
64
+ "selected_utility_gain_over_base": -0.002369848530118664,
65
+ "selector_regret": 0.39043679422854133,
66
+ "success_selector_gap": 0.2,
67
+ "success_support_gap": 0.09333333333333334,
68
+ "support_gap": 0.23527305972451965,
69
+ "tau": 0.0871618711066511
70
+ },
71
+ "PushCube-v1": {
72
+ "base_success": 0.16666666666666666,
73
+ "base_utility": 0.3924041787783305,
74
+ "coverage": 0.8888888888888888,
75
+ "execute_generated": 0.8888888888888888,
76
+ "fallback_rate": 0.1111111111111111,
77
+ "hidden_chart_oracle_success": 0.8333333333333334,
78
+ "hidden_chart_oracle_utility": 1.7015800600250561,
79
+ "outcome_ptr": 0.7777777777777778,
80
+ "predicted_margin": 0.19786446921058906,
81
+ "proposal_oracle_success": 0.6666666666666666,
82
+ "proposal_oracle_utility": 1.365773618221283,
83
+ "selected_candidate_index": 3.111111111111111,
84
+ "selected_success": 0.2777777777777778,
85
+ "selected_success_gain_over_base": 0.1111111111111111,
86
+ "selected_utility": 0.6066941933499442,
87
+ "selected_utility_gain_over_base": 0.21429001457161373,
88
+ "selector_regret": 0.7590794248713387,
89
+ "success_selector_gap": 0.3888888888888889,
90
+ "success_support_gap": 0.3333333333333333,
91
+ "support_gap": 0.634226381778717,
92
+ "tau": 0.08716187110665125
93
+ },
94
+ "StackCube-v1": {
95
+ "base_success": 0.2727272727272727,
96
+ "base_utility": 0.7763947045261209,
97
+ "coverage": 0.3939393939393939,
98
+ "execute_generated": 0.3939393939393939,
99
+ "fallback_rate": 0.6060606060606061,
100
+ "hidden_chart_oracle_success": 1.0,
101
+ "hidden_chart_oracle_utility": 2.0,
102
+ "outcome_ptr": 0.5757575757575758,
103
+ "predicted_margin": 0.07251964474544452,
104
+ "proposal_oracle_success": 0.45454545454545453,
105
+ "proposal_oracle_utility": 1.0546941955884297,
106
+ "selected_candidate_index": 3.5757575757575757,
107
+ "selected_success": 0.2727272727272727,
108
+ "selected_success_gain_over_base": 0.0,
109
+ "selected_utility": 0.7097618074128123,
110
+ "selected_utility_gain_over_base": -0.06663289711330876,
111
+ "selector_regret": 0.348525373321591,
112
+ "success_selector_gap": 0.18181818181818182,
113
+ "success_support_gap": 0.5454545454545454,
114
+ "support_gap": 0.9453058044115702,
115
+ "tau": 0.08716187110665126
116
+ }
117
+ }
workspace/runs/ctt_base_context_obs_learned_dominance_context_envclip_train_to_test/report.md ADDED
@@ -0,0 +1,17 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ # Learned Dominance-Calibrated CTT Selector
2
+
3
+ Calibration rows: `432`
4
+ Eval rows: `144`
5
+ Selected ridge lambda: `10.0`
6
+ Tau: `0.087162`
7
+ Feature set: `context`
8
+ Target: `utility_margin`
9
+
10
+ The ridge calibrator and threshold are fit on calibration measured rows only. Eval outcomes are used only for reporting.
11
+
12
+ | Split | Coverage | Fallback | Base success | Selected success | Proposal oracle | OutcomePTR | Success support gap | Success selector gap |
13
+ | --- | ---: | ---: | ---: | ---: | ---: | ---: | ---: | ---: |
14
+ | calibration | 0.4514 | 0.5486 | 0.2778 | 0.3218 | 0.4722 | 0.4907 | 0.2315 | 0.2014 |
15
+ | eval | 0.3333 | 0.6667 | 0.2917 | 0.3056 | 0.4792 | 0.5069 | 0.2917 | 0.2292 |
16
+
17
+ This is a selector diagnostic over already measured candidates, not a new rollout.
workspace/runs/ctt_base_context_obs_learned_dominance_context_envclip_train_to_test/split_hash.txt ADDED
@@ -0,0 +1,14 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "calibration_target_index": {
3
+ "content_hash": "67f7b4a692f7c7e71da24378dc71e3399c5d35afc14c494869eba6087b765c42",
4
+ "retrieval_index_allowed": true,
5
+ "split": "train",
6
+ "split_hash": "867a94a1c67686c082765671a7bf8d74a9d633085e1b2619c83197c60bf58372"
7
+ },
8
+ "eval_target_index": {
9
+ "content_hash": "0c38d73dc9bc4ab756e12006671b5d12f1a23358b281e54b3f6a9ae3cbf6e9af",
10
+ "retrieval_index_allowed": false,
11
+ "split": "test",
12
+ "split_hash": "84151df9e88c219b76751677558cc5bbbc19082bf358cea7330158e9e403e25f"
13
+ }
14
+ }
workspace/runs/ctt_base_context_obs_learned_dominance_context_envclip_train_to_test/table.tex ADDED
@@ -0,0 +1,8 @@
 
 
 
 
 
 
 
 
 
1
+ % Auto-generated by scripts/eval_learned_dominance_selector.py
2
+ \begin{tabular}{lrrrrrrrr}
3
+ \toprule
4
+ Rows & Coverage & Fallback & Base succ. & Selected succ. & Oracle succ. & OutcomePTR & Succ. support gap & Succ. selector gap \\
5
+ \midrule
6
+ 144 & 0.3333 & 0.6667 & 0.2917 & 0.3056 & 0.4792 & 0.5069 & 0.2917 & 0.2292 \\
7
+ \bottomrule
8
+ \end{tabular}
workspace/runs/ctt_base_context_obs_learned_dominance_context_envclip_train_to_test/train.log ADDED
@@ -0,0 +1,4 @@
 
 
 
 
 
1
+ trained ridge dominance calibrator on calibration measured rows only
2
+ selected_lambda=10.0
3
+ tau=0.087162
4
+ calibration_input=runs/ctt_base_context_obs_train_cal_envclip_rollout_comparison/combined_measured_candidates.json
workspace/runs/ctt_base_context_obs_learned_dominance_context_success_envclip_train_to_test/command.txt ADDED
@@ -0,0 +1 @@
 
 
1
+ python scripts/eval_learned_dominance_selector.py --calibration-input runs/ctt_base_context_obs_train_cal_envclip_rollout_comparison/combined_measured_candidates.json --calibration-target-index data/cil_charts_rgb_refs/train/index.json --eval-input runs/ctt_base_context_obs_test_envclip_rollout_comparison/combined_measured_candidates.json --eval-target-index data/cil_charts_rgb_refs/test/index.json --checkpoint-template runs/ctt_residual_base_context_obs_seed{seed}/model.pt --out-dir runs/ctt_base_context_obs_learned_dominance_context_success_envclip_train_to_test --feature-set context --target success --k 8 --bootstrap-samples 1000
workspace/runs/ctt_base_context_obs_learned_dominance_context_success_envclip_train_to_test/config.yaml ADDED
@@ -0,0 +1,11 @@
 
 
 
 
 
 
 
 
 
 
 
 
1
+ bootstrap_samples: 1000
2
+ calibration_input: runs/ctt_base_context_obs_train_cal_envclip_rollout_comparison/combined_measured_candidates.json
3
+ calibration_target_index: data/cil_charts_rgb_refs/train/index.json
4
+ checkpoint_template: runs/ctt_residual_base_context_obs_seed{seed}/model.pt
5
+ eval_input: runs/ctt_base_context_obs_test_envclip_rollout_comparison/combined_measured_candidates.json
6
+ eval_target_index: data/cil_charts_rgb_refs/test/index.json
7
+ feature_set: context
8
+ k: 8
9
+ out_dir: runs/ctt_base_context_obs_learned_dominance_context_success_envclip_train_to_test
10
+ ridge_lambdas: 0,0.01,0.1,1,10,100
11
+ target: success
workspace/runs/ctt_base_context_obs_learned_dominance_context_success_envclip_train_to_test/data_hash.txt ADDED
@@ -0,0 +1,6 @@
 
 
 
 
 
 
 
1
+ {
2
+ "calibration_input": "06375099001e740a377c402e34b77bf6974be3dc2a468974340ce41eb1b487f0",
3
+ "calibration_target_index": "67973c50072aa87a3b97533b6292b08a8b4d88a95571dc22e91b6e8fd19e0d85",
4
+ "eval_input": "57df43ff4802379dfd60b5209d5a04028d721256406685decededccdbec0d49c",
5
+ "eval_target_index": "dee1b5225485df6624747dd4c5074bcf514685b54017d05e698d2051638fbf36"
6
+ }
workspace/runs/ctt_base_context_obs_learned_dominance_context_success_envclip_train_to_test/eval.log ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ evaluated learned fallback rule on held-out measured rollout rows
2
+ eval_input=runs/ctt_base_context_obs_test_envclip_rollout_comparison/combined_measured_candidates.json
3
+ num_eval_rows=144
workspace/runs/ctt_base_context_obs_learned_dominance_context_success_envclip_train_to_test/git_hash.txt ADDED
@@ -0,0 +1 @@
 
 
1
+ 14e81292e2004925f81e0bb4810f6f144bfd87b7
workspace/runs/ctt_base_context_obs_learned_dominance_context_success_envclip_train_to_test/metrics.json ADDED
The diff for this file is too large to render. See raw diff
 
workspace/runs/ctt_base_context_obs_learned_dominance_context_success_envclip_train_to_test/metrics_by_seed.json ADDED
@@ -0,0 +1,2762 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "1073161778/train0": {
3
+ "base_success": 1.0,
4
+ "base_utility": 2.0,
5
+ "coverage": 0.0,
6
+ "execute_generated": 0.0,
7
+ "fallback_rate": 1.0,
8
+ "hidden_chart_oracle_success": 1.0,
9
+ "hidden_chart_oracle_utility": 2.0,
10
+ "outcome_ptr": 0.0,
11
+ "predicted_margin": 0.27351737405787574,
12
+ "proposal_oracle_success": 1.0,
13
+ "proposal_oracle_utility": 2.0,
14
+ "selected_candidate_index": 1.0,
15
+ "selected_success": 1.0,
16
+ "selected_success_gain_over_base": 0.0,
17
+ "selected_utility": 2.0,
18
+ "selected_utility_gain_over_base": 0.0,
19
+ "selector_regret": 0.0,
20
+ "success_selector_gap": 0.0,
21
+ "success_support_gap": 0.0,
22
+ "support_gap": 0.0,
23
+ "tau": 0.27486681252026457
24
+ },
25
+ "1073161778/train1": {
26
+ "base_success": 1.0,
27
+ "base_utility": 2.0,
28
+ "coverage": 0.0,
29
+ "execute_generated": 0.0,
30
+ "fallback_rate": 1.0,
31
+ "hidden_chart_oracle_success": 1.0,
32
+ "hidden_chart_oracle_utility": 2.0,
33
+ "outcome_ptr": 0.0,
34
+ "predicted_margin": 0.2481774109026027,
35
+ "proposal_oracle_success": 0.0,
36
+ "proposal_oracle_utility": 0.7071996331214905,
37
+ "selected_candidate_index": 5.0,
38
+ "selected_success": 1.0,
39
+ "selected_success_gain_over_base": 0.0,
40
+ "selected_utility": 2.0,
41
+ "selected_utility_gain_over_base": 0.0,
42
+ "selector_regret": 0.0,
43
+ "success_selector_gap": 0.0,
44
+ "success_support_gap": 1.0,
45
+ "support_gap": 1.2928003668785095,
46
+ "tau": 0.27486681252026457
47
+ },
48
+ "1073161778/train2": {
49
+ "base_success": 1.0,
50
+ "base_utility": 2.0,
51
+ "coverage": 0.0,
52
+ "execute_generated": 0.0,
53
+ "fallback_rate": 1.0,
54
+ "hidden_chart_oracle_success": 1.0,
55
+ "hidden_chart_oracle_utility": 2.0,
56
+ "outcome_ptr": 0.0,
57
+ "predicted_margin": 0.25263698343719576,
58
+ "proposal_oracle_success": 1.0,
59
+ "proposal_oracle_utility": 2.0,
60
+ "selected_candidate_index": 3.0,
61
+ "selected_success": 1.0,
62
+ "selected_success_gain_over_base": 0.0,
63
+ "selected_utility": 2.0,
64
+ "selected_utility_gain_over_base": 0.0,
65
+ "selector_regret": 0.0,
66
+ "success_selector_gap": 0.0,
67
+ "success_support_gap": 0.0,
68
+ "support_gap": 0.0,
69
+ "tau": 0.27486681252026457
70
+ },
71
+ "1113371261/train0": {
72
+ "base_success": 0.0,
73
+ "base_utility": 0.15842342376708984,
74
+ "coverage": 0.0,
75
+ "execute_generated": 0.0,
76
+ "fallback_rate": 1.0,
77
+ "hidden_chart_oracle_success": 1.0,
78
+ "hidden_chart_oracle_utility": 2.0,
79
+ "outcome_ptr": 0.0,
80
+ "predicted_margin": 0.26956833152333576,
81
+ "proposal_oracle_success": 0.0,
82
+ "proposal_oracle_utility": 0.10274861752986908,
83
+ "selected_candidate_index": 2.0,
84
+ "selected_success": 0.0,
85
+ "selected_success_gain_over_base": 0.0,
86
+ "selected_utility": 0.15842342376708984,
87
+ "selected_utility_gain_over_base": 0.0,
88
+ "selector_regret": 0.0,
89
+ "success_selector_gap": 0.0,
90
+ "success_support_gap": 1.0,
91
+ "support_gap": 1.897251382470131,
92
+ "tau": 0.27486681252026457
93
+ },
94
+ "1113371261/train1": {
95
+ "base_success": 0.0,
96
+ "base_utility": 0.15842342376708984,
97
+ "coverage": 0.0,
98
+ "execute_generated": 0.0,
99
+ "fallback_rate": 1.0,
100
+ "hidden_chart_oracle_success": 1.0,
101
+ "hidden_chart_oracle_utility": 2.0,
102
+ "outcome_ptr": 1.0,
103
+ "predicted_margin": 0.23570927763802385,
104
+ "proposal_oracle_success": 0.0,
105
+ "proposal_oracle_utility": 0.18093222379684448,
106
+ "selected_candidate_index": 3.0,
107
+ "selected_success": 0.0,
108
+ "selected_success_gain_over_base": 0.0,
109
+ "selected_utility": 0.15842342376708984,
110
+ "selected_utility_gain_over_base": 0.0,
111
+ "selector_regret": 0.02250880002975464,
112
+ "success_selector_gap": 0.0,
113
+ "success_support_gap": 1.0,
114
+ "support_gap": 1.8190677762031555,
115
+ "tau": 0.27486681252026457
116
+ },
117
+ "1113371261/train2": {
118
+ "base_success": 0.0,
119
+ "base_utility": 0.15842342376708984,
120
+ "coverage": 0.0,
121
+ "execute_generated": 0.0,
122
+ "fallback_rate": 1.0,
123
+ "hidden_chart_oracle_success": 1.0,
124
+ "hidden_chart_oracle_utility": 2.0,
125
+ "outcome_ptr": 1.0,
126
+ "predicted_margin": 0.2411081969315667,
127
+ "proposal_oracle_success": 0.0,
128
+ "proposal_oracle_utility": 0.19527167081832886,
129
+ "selected_candidate_index": 5.0,
130
+ "selected_success": 0.0,
131
+ "selected_success_gain_over_base": 0.0,
132
+ "selected_utility": 0.15842342376708984,
133
+ "selected_utility_gain_over_base": 0.0,
134
+ "selector_regret": 0.036848247051239014,
135
+ "success_selector_gap": 0.0,
136
+ "success_support_gap": 1.0,
137
+ "support_gap": 1.8047283291816711,
138
+ "tau": 0.27486681252026457
139
+ },
140
+ "1151410346/train0": {
141
+ "base_success": 0.0,
142
+ "base_utility": 0.2260012924671173,
143
+ "coverage": 1.0,
144
+ "execute_generated": 1.0,
145
+ "fallback_rate": 0.0,
146
+ "hidden_chart_oracle_success": 1.0,
147
+ "hidden_chart_oracle_utility": 2.0,
148
+ "outcome_ptr": 1.0,
149
+ "predicted_margin": 0.36324262482774883,
150
+ "proposal_oracle_success": 0.0,
151
+ "proposal_oracle_utility": 0.8663521409034729,
152
+ "selected_candidate_index": 3.0,
153
+ "selected_success": 0.0,
154
+ "selected_success_gain_over_base": 0.0,
155
+ "selected_utility": 0.22233523428440094,
156
+ "selected_utility_gain_over_base": -0.0036660581827163696,
157
+ "selector_regret": 0.644016906619072,
158
+ "success_selector_gap": 0.0,
159
+ "success_support_gap": 1.0,
160
+ "support_gap": 1.133647859096527,
161
+ "tau": 0.27486681252026457
162
+ },
163
+ "1151410346/train1": {
164
+ "base_success": 0.0,
165
+ "base_utility": 0.2260012924671173,
166
+ "coverage": 1.0,
167
+ "execute_generated": 1.0,
168
+ "fallback_rate": 0.0,
169
+ "hidden_chart_oracle_success": 1.0,
170
+ "hidden_chart_oracle_utility": 2.0,
171
+ "outcome_ptr": 1.0,
172
+ "predicted_margin": 0.2968771702835237,
173
+ "proposal_oracle_success": 0.0,
174
+ "proposal_oracle_utility": 0.8138919472694397,
175
+ "selected_candidate_index": 4.0,
176
+ "selected_success": 0.0,
177
+ "selected_success_gain_over_base": 0.0,
178
+ "selected_utility": 0.8138919472694397,
179
+ "selected_utility_gain_over_base": 0.5878906548023224,
180
+ "selector_regret": 0.0,
181
+ "success_selector_gap": 0.0,
182
+ "success_support_gap": 1.0,
183
+ "support_gap": 1.1861080527305603,
184
+ "tau": 0.27486681252026457
185
+ },
186
+ "1151410346/train2": {
187
+ "base_success": 0.0,
188
+ "base_utility": 0.2260012924671173,
189
+ "coverage": 1.0,
190
+ "execute_generated": 1.0,
191
+ "fallback_rate": 0.0,
192
+ "hidden_chart_oracle_success": 1.0,
193
+ "hidden_chart_oracle_utility": 2.0,
194
+ "outcome_ptr": 0.0,
195
+ "predicted_margin": 0.3232372687485989,
196
+ "proposal_oracle_success": 0.0,
197
+ "proposal_oracle_utility": 0.21912741661071777,
198
+ "selected_candidate_index": 3.0,
199
+ "selected_success": 0.0,
200
+ "selected_success_gain_over_base": 0.0,
201
+ "selected_utility": 0.2153169959783554,
202
+ "selected_utility_gain_over_base": -0.010684296488761902,
203
+ "selector_regret": 0.0038104206323623657,
204
+ "success_selector_gap": 0.0,
205
+ "success_support_gap": 1.0,
206
+ "support_gap": 1.7808725833892822,
207
+ "tau": 0.27486681252026457
208
+ },
209
+ "1261930456/train0": {
210
+ "base_success": 1.0,
211
+ "base_utility": 2.0,
212
+ "coverage": 1.0,
213
+ "execute_generated": 1.0,
214
+ "fallback_rate": 0.0,
215
+ "hidden_chart_oracle_success": 1.0,
216
+ "hidden_chart_oracle_utility": 2.0,
217
+ "outcome_ptr": 0.0,
218
+ "predicted_margin": 0.5151487635357731,
219
+ "proposal_oracle_success": 1.0,
220
+ "proposal_oracle_utility": 2.0,
221
+ "selected_candidate_index": 1.0,
222
+ "selected_success": 1.0,
223
+ "selected_success_gain_over_base": 0.0,
224
+ "selected_utility": 2.0,
225
+ "selected_utility_gain_over_base": 0.0,
226
+ "selector_regret": 0.0,
227
+ "success_selector_gap": 0.0,
228
+ "success_support_gap": 0.0,
229
+ "support_gap": 0.0,
230
+ "tau": 0.27486681252026457
231
+ },
232
+ "1261930456/train1": {
233
+ "base_success": 1.0,
234
+ "base_utility": 2.0,
235
+ "coverage": 1.0,
236
+ "execute_generated": 1.0,
237
+ "fallback_rate": 0.0,
238
+ "hidden_chart_oracle_success": 1.0,
239
+ "hidden_chart_oracle_utility": 2.0,
240
+ "outcome_ptr": 0.0,
241
+ "predicted_margin": 0.47144344762143203,
242
+ "proposal_oracle_success": 1.0,
243
+ "proposal_oracle_utility": 2.0,
244
+ "selected_candidate_index": 4.0,
245
+ "selected_success": 1.0,
246
+ "selected_success_gain_over_base": 0.0,
247
+ "selected_utility": 2.0,
248
+ "selected_utility_gain_over_base": 0.0,
249
+ "selector_regret": 0.0,
250
+ "success_selector_gap": 0.0,
251
+ "success_support_gap": 0.0,
252
+ "support_gap": 0.0,
253
+ "tau": 0.27486681252026457
254
+ },
255
+ "1261930456/train2": {
256
+ "base_success": 1.0,
257
+ "base_utility": 2.0,
258
+ "coverage": 1.0,
259
+ "execute_generated": 1.0,
260
+ "fallback_rate": 0.0,
261
+ "hidden_chart_oracle_success": 1.0,
262
+ "hidden_chart_oracle_utility": 2.0,
263
+ "outcome_ptr": 0.0,
264
+ "predicted_margin": 0.5006602989510598,
265
+ "proposal_oracle_success": 1.0,
266
+ "proposal_oracle_utility": 2.0,
267
+ "selected_candidate_index": 3.0,
268
+ "selected_success": 1.0,
269
+ "selected_success_gain_over_base": 0.0,
270
+ "selected_utility": 2.0,
271
+ "selected_utility_gain_over_base": 0.0,
272
+ "selector_regret": 0.0,
273
+ "success_selector_gap": 0.0,
274
+ "success_support_gap": 0.0,
275
+ "support_gap": 0.0,
276
+ "tau": 0.27486681252026457
277
+ },
278
+ "1324133122/train0": {
279
+ "base_success": 0.0,
280
+ "base_utility": 0.09220491349697113,
281
+ "coverage": 0.0,
282
+ "execute_generated": 0.0,
283
+ "fallback_rate": 1.0,
284
+ "hidden_chart_oracle_success": 1.0,
285
+ "hidden_chart_oracle_utility": 2.0,
286
+ "outcome_ptr": 0.0,
287
+ "predicted_margin": 0.2551089193435261,
288
+ "proposal_oracle_success": 0.0,
289
+ "proposal_oracle_utility": 0.054989203810691833,
290
+ "selected_candidate_index": 3.0,
291
+ "selected_success": 0.0,
292
+ "selected_success_gain_over_base": 0.0,
293
+ "selected_utility": 0.09220491349697113,
294
+ "selected_utility_gain_over_base": 0.0,
295
+ "selector_regret": 0.0,
296
+ "success_selector_gap": 0.0,
297
+ "success_support_gap": 1.0,
298
+ "support_gap": 1.9450107961893082,
299
+ "tau": 0.27486681252026457
300
+ },
301
+ "1324133122/train1": {
302
+ "base_success": 0.0,
303
+ "base_utility": 0.09220491349697113,
304
+ "coverage": 1.0,
305
+ "execute_generated": 1.0,
306
+ "fallback_rate": 0.0,
307
+ "hidden_chart_oracle_success": 1.0,
308
+ "hidden_chart_oracle_utility": 2.0,
309
+ "outcome_ptr": 1.0,
310
+ "predicted_margin": 0.28646037222489074,
311
+ "proposal_oracle_success": 0.0,
312
+ "proposal_oracle_utility": 0.1324586570262909,
313
+ "selected_candidate_index": 1.0,
314
+ "selected_success": 0.0,
315
+ "selected_success_gain_over_base": 0.0,
316
+ "selected_utility": 0.06454251706600189,
317
+ "selected_utility_gain_over_base": -0.02766239643096924,
318
+ "selector_regret": 0.067916139960289,
319
+ "success_selector_gap": 0.0,
320
+ "success_support_gap": 1.0,
321
+ "support_gap": 1.867541342973709,
322
+ "tau": 0.27486681252026457
323
+ },
324
+ "1324133122/train2": {
325
+ "base_success": 0.0,
326
+ "base_utility": 0.09220491349697113,
327
+ "coverage": 0.0,
328
+ "execute_generated": 0.0,
329
+ "fallback_rate": 1.0,
330
+ "hidden_chart_oracle_success": 1.0,
331
+ "hidden_chart_oracle_utility": 2.0,
332
+ "outcome_ptr": 1.0,
333
+ "predicted_margin": 0.22927469986798302,
334
+ "proposal_oracle_success": 0.0,
335
+ "proposal_oracle_utility": 0.19291165471076965,
336
+ "selected_candidate_index": 4.0,
337
+ "selected_success": 0.0,
338
+ "selected_success_gain_over_base": 0.0,
339
+ "selected_utility": 0.09220491349697113,
340
+ "selected_utility_gain_over_base": 0.0,
341
+ "selector_regret": 0.10070674121379852,
342
+ "success_selector_gap": 0.0,
343
+ "success_support_gap": 1.0,
344
+ "support_gap": 1.8070883452892303,
345
+ "tau": 0.27486681252026457
346
+ },
347
+ "1324642411/train0": {
348
+ "base_success": 0.0,
349
+ "base_utility": 0.10647052526473999,
350
+ "coverage": 1.0,
351
+ "execute_generated": 1.0,
352
+ "fallback_rate": 0.0,
353
+ "hidden_chart_oracle_success": 1.0,
354
+ "hidden_chart_oracle_utility": 2.0,
355
+ "outcome_ptr": 1.0,
356
+ "predicted_margin": 0.44544444373667946,
357
+ "proposal_oracle_success": 0.0,
358
+ "proposal_oracle_utility": 0.17193320393562317,
359
+ "selected_candidate_index": 3.0,
360
+ "selected_success": 0.0,
361
+ "selected_success_gain_over_base": 0.0,
362
+ "selected_utility": 0.08104583621025085,
363
+ "selected_utility_gain_over_base": -0.025424689054489136,
364
+ "selector_regret": 0.09088736772537231,
365
+ "success_selector_gap": 0.0,
366
+ "success_support_gap": 1.0,
367
+ "support_gap": 1.8280667960643768,
368
+ "tau": 0.27486681252026457
369
+ },
370
+ "1324642411/train1": {
371
+ "base_success": 0.0,
372
+ "base_utility": 0.10647052526473999,
373
+ "coverage": 1.0,
374
+ "execute_generated": 1.0,
375
+ "fallback_rate": 0.0,
376
+ "hidden_chart_oracle_success": 1.0,
377
+ "hidden_chart_oracle_utility": 2.0,
378
+ "outcome_ptr": 1.0,
379
+ "predicted_margin": 0.514331463098269,
380
+ "proposal_oracle_success": 0.0,
381
+ "proposal_oracle_utility": 0.1836909055709839,
382
+ "selected_candidate_index": 1.0,
383
+ "selected_success": 0.0,
384
+ "selected_success_gain_over_base": 0.0,
385
+ "selected_utility": 0.10393516719341278,
386
+ "selected_utility_gain_over_base": -0.0025353580713272095,
387
+ "selector_regret": 0.0797557383775711,
388
+ "success_selector_gap": 0.0,
389
+ "success_support_gap": 1.0,
390
+ "support_gap": 1.8163090944290161,
391
+ "tau": 0.27486681252026457
392
+ },
393
+ "1324642411/train2": {
394
+ "base_success": 0.0,
395
+ "base_utility": 0.10647052526473999,
396
+ "coverage": 1.0,
397
+ "execute_generated": 1.0,
398
+ "fallback_rate": 0.0,
399
+ "hidden_chart_oracle_success": 1.0,
400
+ "hidden_chart_oracle_utility": 2.0,
401
+ "outcome_ptr": 1.0,
402
+ "predicted_margin": 0.4528383158065934,
403
+ "proposal_oracle_success": 1.0,
404
+ "proposal_oracle_utility": 2.0,
405
+ "selected_candidate_index": 1.0,
406
+ "selected_success": 1.0,
407
+ "selected_success_gain_over_base": 1.0,
408
+ "selected_utility": 2.0,
409
+ "selected_utility_gain_over_base": 1.89352947473526,
410
+ "selector_regret": 0.0,
411
+ "success_selector_gap": 0.0,
412
+ "success_support_gap": 0.0,
413
+ "support_gap": 0.0,
414
+ "tau": 0.27486681252026457
415
+ },
416
+ "1375492786/train0": {
417
+ "base_success": 1.0,
418
+ "base_utility": 2.0,
419
+ "coverage": 0.0,
420
+ "execute_generated": 0.0,
421
+ "fallback_rate": 1.0,
422
+ "hidden_chart_oracle_success": 1.0,
423
+ "hidden_chart_oracle_utility": 2.0,
424
+ "outcome_ptr": 0.0,
425
+ "predicted_margin": 0.22163999599382528,
426
+ "proposal_oracle_success": 1.0,
427
+ "proposal_oracle_utility": 2.0,
428
+ "selected_candidate_index": 2.0,
429
+ "selected_success": 1.0,
430
+ "selected_success_gain_over_base": 0.0,
431
+ "selected_utility": 2.0,
432
+ "selected_utility_gain_over_base": 0.0,
433
+ "selector_regret": 0.0,
434
+ "success_selector_gap": 0.0,
435
+ "success_support_gap": 0.0,
436
+ "support_gap": 0.0,
437
+ "tau": 0.27486681252026457
438
+ },
439
+ "1375492786/train1": {
440
+ "base_success": 1.0,
441
+ "base_utility": 2.0,
442
+ "coverage": 0.0,
443
+ "execute_generated": 0.0,
444
+ "fallback_rate": 1.0,
445
+ "hidden_chart_oracle_success": 1.0,
446
+ "hidden_chart_oracle_utility": 2.0,
447
+ "outcome_ptr": 0.0,
448
+ "predicted_margin": 0.21743073433645888,
449
+ "proposal_oracle_success": 0.0,
450
+ "proposal_oracle_utility": 0.08251869678497314,
451
+ "selected_candidate_index": 3.0,
452
+ "selected_success": 1.0,
453
+ "selected_success_gain_over_base": 0.0,
454
+ "selected_utility": 2.0,
455
+ "selected_utility_gain_over_base": 0.0,
456
+ "selector_regret": 0.0,
457
+ "success_selector_gap": 0.0,
458
+ "success_support_gap": 1.0,
459
+ "support_gap": 1.9174813032150269,
460
+ "tau": 0.27486681252026457
461
+ },
462
+ "1375492786/train2": {
463
+ "base_success": 1.0,
464
+ "base_utility": 2.0,
465
+ "coverage": 0.0,
466
+ "execute_generated": 0.0,
467
+ "fallback_rate": 1.0,
468
+ "hidden_chart_oracle_success": 1.0,
469
+ "hidden_chart_oracle_utility": 2.0,
470
+ "outcome_ptr": 0.0,
471
+ "predicted_margin": 0.1567293104998793,
472
+ "proposal_oracle_success": 1.0,
473
+ "proposal_oracle_utility": 2.0,
474
+ "selected_candidate_index": 2.0,
475
+ "selected_success": 1.0,
476
+ "selected_success_gain_over_base": 0.0,
477
+ "selected_utility": 2.0,
478
+ "selected_utility_gain_over_base": 0.0,
479
+ "selector_regret": 0.0,
480
+ "success_selector_gap": 0.0,
481
+ "success_support_gap": 0.0,
482
+ "support_gap": 0.0,
483
+ "tau": 0.27486681252026457
484
+ },
485
+ "141984615/train0": {
486
+ "base_success": 0.0,
487
+ "base_utility": 0.10283424705266953,
488
+ "coverage": 0.0,
489
+ "execute_generated": 0.0,
490
+ "fallback_rate": 1.0,
491
+ "hidden_chart_oracle_success": 0.0,
492
+ "hidden_chart_oracle_utility": 0.2857772707939148,
493
+ "outcome_ptr": 0.0,
494
+ "predicted_margin": 0.25906691818362704,
495
+ "proposal_oracle_success": 0.0,
496
+ "proposal_oracle_utility": 0.007291515823453665,
497
+ "selected_candidate_index": 7.0,
498
+ "selected_success": 0.0,
499
+ "selected_success_gain_over_base": 0.0,
500
+ "selected_utility": 0.10283424705266953,
501
+ "selected_utility_gain_over_base": 0.0,
502
+ "selector_regret": 0.0,
503
+ "success_selector_gap": 0.0,
504
+ "success_support_gap": 0.0,
505
+ "support_gap": 0.27848575497046113,
506
+ "tau": 0.27486681252026457
507
+ },
508
+ "141984615/train1": {
509
+ "base_success": 0.0,
510
+ "base_utility": 0.10283424705266953,
511
+ "coverage": 1.0,
512
+ "execute_generated": 1.0,
513
+ "fallback_rate": 0.0,
514
+ "hidden_chart_oracle_success": 0.0,
515
+ "hidden_chart_oracle_utility": 0.2857772707939148,
516
+ "outcome_ptr": 0.0,
517
+ "predicted_margin": 0.31966486250851045,
518
+ "proposal_oracle_success": 0.0,
519
+ "proposal_oracle_utility": 0.037011127918958664,
520
+ "selected_candidate_index": 1.0,
521
+ "selected_success": 0.0,
522
+ "selected_success_gain_over_base": 0.0,
523
+ "selected_utility": 0.027254482731223106,
524
+ "selected_utility_gain_over_base": -0.07557976432144642,
525
+ "selector_regret": 0.009756645187735558,
526
+ "success_selector_gap": 0.0,
527
+ "success_support_gap": 0.0,
528
+ "support_gap": 0.24876614287495613,
529
+ "tau": 0.27486681252026457
530
+ },
531
+ "141984615/train2": {
532
+ "base_success": 0.0,
533
+ "base_utility": 0.10283424705266953,
534
+ "coverage": 1.0,
535
+ "execute_generated": 1.0,
536
+ "fallback_rate": 0.0,
537
+ "hidden_chart_oracle_success": 0.0,
538
+ "hidden_chart_oracle_utility": 0.2857772707939148,
539
+ "outcome_ptr": 0.0,
540
+ "predicted_margin": 0.3355557686554944,
541
+ "proposal_oracle_success": 0.0,
542
+ "proposal_oracle_utility": 0.030228236690163612,
543
+ "selected_candidate_index": 6.0,
544
+ "selected_success": 0.0,
545
+ "selected_success_gain_over_base": 0.0,
546
+ "selected_utility": 0.020666897296905518,
547
+ "selected_utility_gain_over_base": -0.08216734975576401,
548
+ "selector_regret": 0.009561339393258095,
549
+ "success_selector_gap": 0.0,
550
+ "success_support_gap": 0.0,
551
+ "support_gap": 0.2555490341037512,
552
+ "tau": 0.27486681252026457
553
+ },
554
+ "1461069713/train0": {
555
+ "base_success": 0.0,
556
+ "base_utility": 0.0939030647277832,
557
+ "coverage": 1.0,
558
+ "execute_generated": 1.0,
559
+ "fallback_rate": 0.0,
560
+ "hidden_chart_oracle_success": 0.0,
561
+ "hidden_chart_oracle_utility": 0.1593424528837204,
562
+ "outcome_ptr": 0.0,
563
+ "predicted_margin": 0.36227395129135637,
564
+ "proposal_oracle_success": 0.0,
565
+ "proposal_oracle_utility": 0.06019836664199829,
566
+ "selected_candidate_index": 3.0,
567
+ "selected_success": 0.0,
568
+ "selected_success_gain_over_base": 0.0,
569
+ "selected_utility": 0.03448672965168953,
570
+ "selected_utility_gain_over_base": -0.059416335076093674,
571
+ "selector_regret": 0.02571163699030876,
572
+ "success_selector_gap": 0.0,
573
+ "success_support_gap": 0.0,
574
+ "support_gap": 0.0991440862417221,
575
+ "tau": 0.27486681252026457
576
+ },
577
+ "1461069713/train1": {
578
+ "base_success": 0.0,
579
+ "base_utility": 0.0939030647277832,
580
+ "coverage": 0.0,
581
+ "execute_generated": 0.0,
582
+ "fallback_rate": 1.0,
583
+ "hidden_chart_oracle_success": 0.0,
584
+ "hidden_chart_oracle_utility": 0.1593424528837204,
585
+ "outcome_ptr": 0.0,
586
+ "predicted_margin": 0.2181904542086462,
587
+ "proposal_oracle_success": 0.0,
588
+ "proposal_oracle_utility": 0.016585012897849083,
589
+ "selected_candidate_index": 1.0,
590
+ "selected_success": 0.0,
591
+ "selected_success_gain_over_base": 0.0,
592
+ "selected_utility": 0.0939030647277832,
593
+ "selected_utility_gain_over_base": 0.0,
594
+ "selector_regret": 0.0,
595
+ "success_selector_gap": 0.0,
596
+ "success_support_gap": 0.0,
597
+ "support_gap": 0.14275743998587132,
598
+ "tau": 0.27486681252026457
599
+ },
600
+ "1461069713/train2": {
601
+ "base_success": 0.0,
602
+ "base_utility": 0.0939030647277832,
603
+ "coverage": 1.0,
604
+ "execute_generated": 1.0,
605
+ "fallback_rate": 0.0,
606
+ "hidden_chart_oracle_success": 0.0,
607
+ "hidden_chart_oracle_utility": 0.1593424528837204,
608
+ "outcome_ptr": 1.0,
609
+ "predicted_margin": 0.4108679110767711,
610
+ "proposal_oracle_success": 0.0,
611
+ "proposal_oracle_utility": 0.09970339387655258,
612
+ "selected_candidate_index": 1.0,
613
+ "selected_success": 0.0,
614
+ "selected_success_gain_over_base": 0.0,
615
+ "selected_utility": 0.09168759733438492,
616
+ "selected_utility_gain_over_base": -0.002215467393398285,
617
+ "selector_regret": 0.008015796542167664,
618
+ "success_selector_gap": 0.0,
619
+ "success_support_gap": 0.0,
620
+ "support_gap": 0.059639059007167816,
621
+ "tau": 0.27486681252026457
622
+ },
623
+ "1480856810/train0": {
624
+ "base_success": 0.0,
625
+ "base_utility": 0.08703407645225525,
626
+ "coverage": 1.0,
627
+ "execute_generated": 1.0,
628
+ "fallback_rate": 0.0,
629
+ "hidden_chart_oracle_success": 0.0,
630
+ "hidden_chart_oracle_utility": 0.20948036015033722,
631
+ "outcome_ptr": 1.0,
632
+ "predicted_margin": 0.5517417835206935,
633
+ "proposal_oracle_success": 1.0,
634
+ "proposal_oracle_utility": 2.0,
635
+ "selected_candidate_index": 1.0,
636
+ "selected_success": 1.0,
637
+ "selected_success_gain_over_base": 1.0,
638
+ "selected_utility": 2.0,
639
+ "selected_utility_gain_over_base": 1.9129659235477448,
640
+ "selector_regret": 0.0,
641
+ "success_selector_gap": 0.0,
642
+ "success_support_gap": 0.0,
643
+ "support_gap": 0.0,
644
+ "tau": 0.27486681252026457
645
+ },
646
+ "1480856810/train1": {
647
+ "base_success": 0.0,
648
+ "base_utility": 0.08703407645225525,
649
+ "coverage": 1.0,
650
+ "execute_generated": 1.0,
651
+ "fallback_rate": 0.0,
652
+ "hidden_chart_oracle_success": 0.0,
653
+ "hidden_chart_oracle_utility": 0.20948036015033722,
654
+ "outcome_ptr": 1.0,
655
+ "predicted_margin": 0.5205859954722791,
656
+ "proposal_oracle_success": 1.0,
657
+ "proposal_oracle_utility": 2.0,
658
+ "selected_candidate_index": 0.0,
659
+ "selected_success": 0.0,
660
+ "selected_success_gain_over_base": 0.0,
661
+ "selected_utility": 0.20867714285850525,
662
+ "selected_utility_gain_over_base": 0.12164306640625,
663
+ "selector_regret": 1.7913228571414948,
664
+ "success_selector_gap": 1.0,
665
+ "success_support_gap": 0.0,
666
+ "support_gap": 0.0,
667
+ "tau": 0.27486681252026457
668
+ },
669
+ "1480856810/train2": {
670
+ "base_success": 0.0,
671
+ "base_utility": 0.08703407645225525,
672
+ "coverage": 1.0,
673
+ "execute_generated": 1.0,
674
+ "fallback_rate": 0.0,
675
+ "hidden_chart_oracle_success": 0.0,
676
+ "hidden_chart_oracle_utility": 0.20948036015033722,
677
+ "outcome_ptr": 1.0,
678
+ "predicted_margin": 0.5153274975271694,
679
+ "proposal_oracle_success": 1.0,
680
+ "proposal_oracle_utility": 2.0,
681
+ "selected_candidate_index": 2.0,
682
+ "selected_success": 0.0,
683
+ "selected_success_gain_over_base": 0.0,
684
+ "selected_utility": 0.21894000470638275,
685
+ "selected_utility_gain_over_base": 0.1319059282541275,
686
+ "selector_regret": 1.7810599952936172,
687
+ "success_selector_gap": 1.0,
688
+ "success_support_gap": 0.0,
689
+ "support_gap": 0.0,
690
+ "tau": 0.27486681252026457
691
+ },
692
+ "1509980664/train0": {
693
+ "base_success": 0.0,
694
+ "base_utility": 0.8334100246429443,
695
+ "coverage": 1.0,
696
+ "execute_generated": 1.0,
697
+ "fallback_rate": 0.0,
698
+ "hidden_chart_oracle_success": 1.0,
699
+ "hidden_chart_oracle_utility": 2.0,
700
+ "outcome_ptr": 1.0,
701
+ "predicted_margin": 0.31798957790509685,
702
+ "proposal_oracle_success": 1.0,
703
+ "proposal_oracle_utility": 2.0,
704
+ "selected_candidate_index": 1.0,
705
+ "selected_success": 0.0,
706
+ "selected_success_gain_over_base": 0.0,
707
+ "selected_utility": 0.1792946457862854,
708
+ "selected_utility_gain_over_base": -0.6541153788566589,
709
+ "selector_regret": 1.8207053542137146,
710
+ "success_selector_gap": 1.0,
711
+ "success_support_gap": 0.0,
712
+ "support_gap": 0.0,
713
+ "tau": 0.27486681252026457
714
+ },
715
+ "1509980664/train1": {
716
+ "base_success": 0.0,
717
+ "base_utility": 0.8334100246429443,
718
+ "coverage": 1.0,
719
+ "execute_generated": 1.0,
720
+ "fallback_rate": 0.0,
721
+ "hidden_chart_oracle_success": 1.0,
722
+ "hidden_chart_oracle_utility": 2.0,
723
+ "outcome_ptr": 1.0,
724
+ "predicted_margin": 0.32004241670590144,
725
+ "proposal_oracle_success": 1.0,
726
+ "proposal_oracle_utility": 2.0,
727
+ "selected_candidate_index": 4.0,
728
+ "selected_success": 0.0,
729
+ "selected_success_gain_over_base": 0.0,
730
+ "selected_utility": 0.21694543957710266,
731
+ "selected_utility_gain_over_base": -0.6164645850658417,
732
+ "selector_regret": 1.7830545604228973,
733
+ "success_selector_gap": 1.0,
734
+ "success_support_gap": 0.0,
735
+ "support_gap": 0.0,
736
+ "tau": 0.27486681252026457
737
+ },
738
+ "1509980664/train2": {
739
+ "base_success": 0.0,
740
+ "base_utility": 0.8334100246429443,
741
+ "coverage": 1.0,
742
+ "execute_generated": 1.0,
743
+ "fallback_rate": 0.0,
744
+ "hidden_chart_oracle_success": 1.0,
745
+ "hidden_chart_oracle_utility": 2.0,
746
+ "outcome_ptr": 0.0,
747
+ "predicted_margin": 0.30322779688903584,
748
+ "proposal_oracle_success": 0.0,
749
+ "proposal_oracle_utility": 0.8162217140197754,
750
+ "selected_candidate_index": 3.0,
751
+ "selected_success": 0.0,
752
+ "selected_success_gain_over_base": 0.0,
753
+ "selected_utility": 0.18445390462875366,
754
+ "selected_utility_gain_over_base": -0.6489561200141907,
755
+ "selector_regret": 0.6317678093910217,
756
+ "success_selector_gap": 0.0,
757
+ "success_support_gap": 1.0,
758
+ "support_gap": 1.1837782859802246,
759
+ "tau": 0.27486681252026457
760
+ },
761
+ "1522272686/train0": {
762
+ "base_success": 0.0,
763
+ "base_utility": 0.5839591026306152,
764
+ "coverage": 1.0,
765
+ "execute_generated": 1.0,
766
+ "fallback_rate": 0.0,
767
+ "hidden_chart_oracle_success": 1.0,
768
+ "hidden_chart_oracle_utility": 2.0,
769
+ "outcome_ptr": 0.0,
770
+ "predicted_margin": 0.3720099739745567,
771
+ "proposal_oracle_success": 0.0,
772
+ "proposal_oracle_utility": 0.18325337767601013,
773
+ "selected_candidate_index": 1.0,
774
+ "selected_success": 0.0,
775
+ "selected_success_gain_over_base": 0.0,
776
+ "selected_utility": 0.08508012443780899,
777
+ "selected_utility_gain_over_base": -0.49887897819280624,
778
+ "selector_regret": 0.09817325323820114,
779
+ "success_selector_gap": 0.0,
780
+ "success_support_gap": 1.0,
781
+ "support_gap": 1.8167466223239899,
782
+ "tau": 0.27486681252026457
783
+ },
784
+ "1522272686/train1": {
785
+ "base_success": 0.0,
786
+ "base_utility": 0.5839591026306152,
787
+ "coverage": 0.0,
788
+ "execute_generated": 0.0,
789
+ "fallback_rate": 1.0,
790
+ "hidden_chart_oracle_success": 1.0,
791
+ "hidden_chart_oracle_utility": 2.0,
792
+ "outcome_ptr": 0.0,
793
+ "predicted_margin": 0.22768582422987366,
794
+ "proposal_oracle_success": 0.0,
795
+ "proposal_oracle_utility": 0.07014383375644684,
796
+ "selected_candidate_index": 1.0,
797
+ "selected_success": 0.0,
798
+ "selected_success_gain_over_base": 0.0,
799
+ "selected_utility": 0.5839591026306152,
800
+ "selected_utility_gain_over_base": 0.0,
801
+ "selector_regret": 0.0,
802
+ "success_selector_gap": 0.0,
803
+ "success_support_gap": 1.0,
804
+ "support_gap": 1.9298561662435532,
805
+ "tau": 0.27486681252026457
806
+ },
807
+ "1522272686/train2": {
808
+ "base_success": 0.0,
809
+ "base_utility": 0.5839591026306152,
810
+ "coverage": 0.0,
811
+ "execute_generated": 0.0,
812
+ "fallback_rate": 1.0,
813
+ "hidden_chart_oracle_success": 1.0,
814
+ "hidden_chart_oracle_utility": 2.0,
815
+ "outcome_ptr": 0.0,
816
+ "predicted_margin": 0.18976407960854358,
817
+ "proposal_oracle_success": 0.0,
818
+ "proposal_oracle_utility": 0.14937306940555573,
819
+ "selected_candidate_index": 4.0,
820
+ "selected_success": 0.0,
821
+ "selected_success_gain_over_base": 0.0,
822
+ "selected_utility": 0.5839591026306152,
823
+ "selected_utility_gain_over_base": 0.0,
824
+ "selector_regret": 0.0,
825
+ "success_selector_gap": 0.0,
826
+ "success_support_gap": 1.0,
827
+ "support_gap": 1.8506269305944443,
828
+ "tau": 0.27486681252026457
829
+ },
830
+ "1549324805/train0": {
831
+ "base_success": 0.0,
832
+ "base_utility": 0.10451483726501465,
833
+ "coverage": 1.0,
834
+ "execute_generated": 1.0,
835
+ "fallback_rate": 0.0,
836
+ "hidden_chart_oracle_success": 0.0,
837
+ "hidden_chart_oracle_utility": 0.2974560856819153,
838
+ "outcome_ptr": 0.0,
839
+ "predicted_margin": 0.41031975368727036,
840
+ "proposal_oracle_success": 0.0,
841
+ "proposal_oracle_utility": 0.0415392704308033,
842
+ "selected_candidate_index": 0.0,
843
+ "selected_success": 0.0,
844
+ "selected_success_gain_over_base": 0.0,
845
+ "selected_utility": 0.0260902252048254,
846
+ "selected_utility_gain_over_base": -0.07842461206018925,
847
+ "selector_regret": 0.015449045225977898,
848
+ "success_selector_gap": 0.0,
849
+ "success_support_gap": 0.0,
850
+ "support_gap": 0.255916815251112,
851
+ "tau": 0.27486681252026457
852
+ },
853
+ "1549324805/train1": {
854
+ "base_success": 0.0,
855
+ "base_utility": 0.10451483726501465,
856
+ "coverage": 1.0,
857
+ "execute_generated": 1.0,
858
+ "fallback_rate": 0.0,
859
+ "hidden_chart_oracle_success": 0.0,
860
+ "hidden_chart_oracle_utility": 0.2974560856819153,
861
+ "outcome_ptr": 1.0,
862
+ "predicted_margin": 0.42814869442608283,
863
+ "proposal_oracle_success": 0.0,
864
+ "proposal_oracle_utility": 0.19892151653766632,
865
+ "selected_candidate_index": 0.0,
866
+ "selected_success": 0.0,
867
+ "selected_success_gain_over_base": 0.0,
868
+ "selected_utility": 0.0303483996540308,
869
+ "selected_utility_gain_over_base": -0.07416643761098385,
870
+ "selector_regret": 0.16857311688363552,
871
+ "success_selector_gap": 0.0,
872
+ "success_support_gap": 0.0,
873
+ "support_gap": 0.09853456914424896,
874
+ "tau": 0.27486681252026457
875
+ },
876
+ "1549324805/train2": {
877
+ "base_success": 0.0,
878
+ "base_utility": 0.10451483726501465,
879
+ "coverage": 1.0,
880
+ "execute_generated": 1.0,
881
+ "fallback_rate": 0.0,
882
+ "hidden_chart_oracle_success": 0.0,
883
+ "hidden_chart_oracle_utility": 0.2974560856819153,
884
+ "outcome_ptr": 0.0,
885
+ "predicted_margin": 0.3815901785915338,
886
+ "proposal_oracle_success": 0.0,
887
+ "proposal_oracle_utility": 0.09092829376459122,
888
+ "selected_candidate_index": 1.0,
889
+ "selected_success": 0.0,
890
+ "selected_success_gain_over_base": 0.0,
891
+ "selected_utility": 0.04690708592534065,
892
+ "selected_utility_gain_over_base": -0.057607751339673996,
893
+ "selector_regret": 0.044021207839250565,
894
+ "success_selector_gap": 0.0,
895
+ "success_support_gap": 0.0,
896
+ "support_gap": 0.20652779191732407,
897
+ "tau": 0.27486681252026457
898
+ },
899
+ "1578990937/train0": {
900
+ "base_success": 0.0,
901
+ "base_utility": 0.82413649559021,
902
+ "coverage": 1.0,
903
+ "execute_generated": 1.0,
904
+ "fallback_rate": 0.0,
905
+ "hidden_chart_oracle_success": 1.0,
906
+ "hidden_chart_oracle_utility": 2.0,
907
+ "outcome_ptr": 1.0,
908
+ "predicted_margin": 0.3536294028807183,
909
+ "proposal_oracle_success": 1.0,
910
+ "proposal_oracle_utility": 2.0,
911
+ "selected_candidate_index": 7.0,
912
+ "selected_success": 0.0,
913
+ "selected_success_gain_over_base": 0.0,
914
+ "selected_utility": 0.16207581758499146,
915
+ "selected_utility_gain_over_base": -0.6620606780052185,
916
+ "selector_regret": 1.8379241824150085,
917
+ "success_selector_gap": 1.0,
918
+ "success_support_gap": 0.0,
919
+ "support_gap": 0.0,
920
+ "tau": 0.27486681252026457
921
+ },
922
+ "1578990937/train1": {
923
+ "base_success": 0.0,
924
+ "base_utility": 0.82413649559021,
925
+ "coverage": 1.0,
926
+ "execute_generated": 1.0,
927
+ "fallback_rate": 0.0,
928
+ "hidden_chart_oracle_success": 1.0,
929
+ "hidden_chart_oracle_utility": 2.0,
930
+ "outcome_ptr": 1.0,
931
+ "predicted_margin": 0.2982680610570334,
932
+ "proposal_oracle_success": 1.0,
933
+ "proposal_oracle_utility": 2.0,
934
+ "selected_candidate_index": 3.0,
935
+ "selected_success": 1.0,
936
+ "selected_success_gain_over_base": 1.0,
937
+ "selected_utility": 2.0,
938
+ "selected_utility_gain_over_base": 1.17586350440979,
939
+ "selector_regret": 0.0,
940
+ "success_selector_gap": 0.0,
941
+ "success_support_gap": 0.0,
942
+ "support_gap": 0.0,
943
+ "tau": 0.27486681252026457
944
+ },
945
+ "1578990937/train2": {
946
+ "base_success": 0.0,
947
+ "base_utility": 0.82413649559021,
948
+ "coverage": 1.0,
949
+ "execute_generated": 1.0,
950
+ "fallback_rate": 0.0,
951
+ "hidden_chart_oracle_success": 1.0,
952
+ "hidden_chart_oracle_utility": 2.0,
953
+ "outcome_ptr": 1.0,
954
+ "predicted_margin": 0.3132319202957881,
955
+ "proposal_oracle_success": 1.0,
956
+ "proposal_oracle_utility": 2.0,
957
+ "selected_candidate_index": 1.0,
958
+ "selected_success": 0.0,
959
+ "selected_success_gain_over_base": 0.0,
960
+ "selected_utility": 0.1460946798324585,
961
+ "selected_utility_gain_over_base": -0.6780418157577515,
962
+ "selector_regret": 1.8539053201675415,
963
+ "success_selector_gap": 1.0,
964
+ "success_support_gap": 0.0,
965
+ "support_gap": 0.0,
966
+ "tau": 0.27486681252026457
967
+ },
968
+ "1747178030/train0": {
969
+ "base_success": 1.0,
970
+ "base_utility": 2.0,
971
+ "coverage": 1.0,
972
+ "execute_generated": 1.0,
973
+ "fallback_rate": 0.0,
974
+ "hidden_chart_oracle_success": 1.0,
975
+ "hidden_chart_oracle_utility": 2.0,
976
+ "outcome_ptr": 0.0,
977
+ "predicted_margin": 0.3444861450293654,
978
+ "proposal_oracle_success": 1.0,
979
+ "proposal_oracle_utility": 2.0,
980
+ "selected_candidate_index": 1.0,
981
+ "selected_success": 1.0,
982
+ "selected_success_gain_over_base": 0.0,
983
+ "selected_utility": 2.0,
984
+ "selected_utility_gain_over_base": 0.0,
985
+ "selector_regret": 0.0,
986
+ "success_selector_gap": 0.0,
987
+ "success_support_gap": 0.0,
988
+ "support_gap": 0.0,
989
+ "tau": 0.27486681252026457
990
+ },
991
+ "1747178030/train1": {
992
+ "base_success": 1.0,
993
+ "base_utility": 2.0,
994
+ "coverage": 0.0,
995
+ "execute_generated": 0.0,
996
+ "fallback_rate": 1.0,
997
+ "hidden_chart_oracle_success": 1.0,
998
+ "hidden_chart_oracle_utility": 2.0,
999
+ "outcome_ptr": 0.0,
1000
+ "predicted_margin": 0.26979063255987656,
1001
+ "proposal_oracle_success": 1.0,
1002
+ "proposal_oracle_utility": 2.0,
1003
+ "selected_candidate_index": 4.0,
1004
+ "selected_success": 1.0,
1005
+ "selected_success_gain_over_base": 0.0,
1006
+ "selected_utility": 2.0,
1007
+ "selected_utility_gain_over_base": 0.0,
1008
+ "selector_regret": 0.0,
1009
+ "success_selector_gap": 0.0,
1010
+ "success_support_gap": 0.0,
1011
+ "support_gap": 0.0,
1012
+ "tau": 0.27486681252026457
1013
+ },
1014
+ "1747178030/train2": {
1015
+ "base_success": 1.0,
1016
+ "base_utility": 2.0,
1017
+ "coverage": 1.0,
1018
+ "execute_generated": 1.0,
1019
+ "fallback_rate": 0.0,
1020
+ "hidden_chart_oracle_success": 1.0,
1021
+ "hidden_chart_oracle_utility": 2.0,
1022
+ "outcome_ptr": 0.0,
1023
+ "predicted_margin": 0.2949234302495328,
1024
+ "proposal_oracle_success": 1.0,
1025
+ "proposal_oracle_utility": 2.0,
1026
+ "selected_candidate_index": 6.0,
1027
+ "selected_success": 1.0,
1028
+ "selected_success_gain_over_base": 0.0,
1029
+ "selected_utility": 2.0,
1030
+ "selected_utility_gain_over_base": 0.0,
1031
+ "selector_regret": 0.0,
1032
+ "success_selector_gap": 0.0,
1033
+ "success_support_gap": 0.0,
1034
+ "support_gap": 0.0,
1035
+ "tau": 0.27486681252026457
1036
+ },
1037
+ "1790188787/train0": {
1038
+ "base_success": 0.0,
1039
+ "base_utility": 0.12476746737957001,
1040
+ "coverage": 1.0,
1041
+ "execute_generated": 1.0,
1042
+ "fallback_rate": 0.0,
1043
+ "hidden_chart_oracle_success": 1.0,
1044
+ "hidden_chart_oracle_utility": 2.0,
1045
+ "outcome_ptr": 0.0,
1046
+ "predicted_margin": 0.488522482876192,
1047
+ "proposal_oracle_success": 0.0,
1048
+ "proposal_oracle_utility": 0.09791924059391022,
1049
+ "selected_candidate_index": 0.0,
1050
+ "selected_success": 0.0,
1051
+ "selected_success_gain_over_base": 0.0,
1052
+ "selected_utility": 0.06870591640472412,
1053
+ "selected_utility_gain_over_base": -0.056061550974845886,
1054
+ "selector_regret": 0.029213324189186096,
1055
+ "success_selector_gap": 0.0,
1056
+ "success_support_gap": 1.0,
1057
+ "support_gap": 1.9020807594060898,
1058
+ "tau": 0.27486681252026457
1059
+ },
1060
+ "1790188787/train1": {
1061
+ "base_success": 0.0,
1062
+ "base_utility": 0.12476746737957001,
1063
+ "coverage": 1.0,
1064
+ "execute_generated": 1.0,
1065
+ "fallback_rate": 0.0,
1066
+ "hidden_chart_oracle_success": 1.0,
1067
+ "hidden_chart_oracle_utility": 2.0,
1068
+ "outcome_ptr": 1.0,
1069
+ "predicted_margin": 0.4664307166911985,
1070
+ "proposal_oracle_success": 0.0,
1071
+ "proposal_oracle_utility": 0.13031470775604248,
1072
+ "selected_candidate_index": 4.0,
1073
+ "selected_success": 0.0,
1074
+ "selected_success_gain_over_base": 0.0,
1075
+ "selected_utility": 0.08738848567008972,
1076
+ "selected_utility_gain_over_base": -0.037378981709480286,
1077
+ "selector_regret": 0.04292622208595276,
1078
+ "success_selector_gap": 0.0,
1079
+ "success_support_gap": 1.0,
1080
+ "support_gap": 1.8696852922439575,
1081
+ "tau": 0.27486681252026457
1082
+ },
1083
+ "1790188787/train2": {
1084
+ "base_success": 0.0,
1085
+ "base_utility": 0.12476746737957001,
1086
+ "coverage": 1.0,
1087
+ "execute_generated": 1.0,
1088
+ "fallback_rate": 0.0,
1089
+ "hidden_chart_oracle_success": 1.0,
1090
+ "hidden_chart_oracle_utility": 2.0,
1091
+ "outcome_ptr": 1.0,
1092
+ "predicted_margin": 0.5144353353506044,
1093
+ "proposal_oracle_success": 1.0,
1094
+ "proposal_oracle_utility": 2.0,
1095
+ "selected_candidate_index": 1.0,
1096
+ "selected_success": 0.0,
1097
+ "selected_success_gain_over_base": 0.0,
1098
+ "selected_utility": 0.1341685652732849,
1099
+ "selected_utility_gain_over_base": 0.009401097893714905,
1100
+ "selector_regret": 1.865831434726715,
1101
+ "success_selector_gap": 1.0,
1102
+ "success_support_gap": 0.0,
1103
+ "support_gap": 0.0,
1104
+ "tau": 0.27486681252026457
1105
+ },
1106
+ "1822210471/train0": {
1107
+ "base_success": 0.0,
1108
+ "base_utility": 0.07191530615091324,
1109
+ "coverage": 0.0,
1110
+ "execute_generated": 0.0,
1111
+ "fallback_rate": 1.0,
1112
+ "hidden_chart_oracle_success": 1.0,
1113
+ "hidden_chart_oracle_utility": 2.0,
1114
+ "outcome_ptr": 1.0,
1115
+ "predicted_margin": 0.236039507547914,
1116
+ "proposal_oracle_success": 1.0,
1117
+ "proposal_oracle_utility": 2.0,
1118
+ "selected_candidate_index": 2.0,
1119
+ "selected_success": 0.0,
1120
+ "selected_success_gain_over_base": 0.0,
1121
+ "selected_utility": 0.07191530615091324,
1122
+ "selected_utility_gain_over_base": 0.0,
1123
+ "selector_regret": 1.9280846938490868,
1124
+ "success_selector_gap": 1.0,
1125
+ "success_support_gap": 0.0,
1126
+ "support_gap": 0.0,
1127
+ "tau": 0.27486681252026457
1128
+ },
1129
+ "1822210471/train1": {
1130
+ "base_success": 0.0,
1131
+ "base_utility": 0.07191530615091324,
1132
+ "coverage": 0.0,
1133
+ "execute_generated": 0.0,
1134
+ "fallback_rate": 1.0,
1135
+ "hidden_chart_oracle_success": 1.0,
1136
+ "hidden_chart_oracle_utility": 2.0,
1137
+ "outcome_ptr": 1.0,
1138
+ "predicted_margin": 0.20275936799419922,
1139
+ "proposal_oracle_success": 1.0,
1140
+ "proposal_oracle_utility": 2.0,
1141
+ "selected_candidate_index": 3.0,
1142
+ "selected_success": 0.0,
1143
+ "selected_success_gain_over_base": 0.0,
1144
+ "selected_utility": 0.07191530615091324,
1145
+ "selected_utility_gain_over_base": 0.0,
1146
+ "selector_regret": 1.9280846938490868,
1147
+ "success_selector_gap": 1.0,
1148
+ "success_support_gap": 0.0,
1149
+ "support_gap": 0.0,
1150
+ "tau": 0.27486681252026457
1151
+ },
1152
+ "1822210471/train2": {
1153
+ "base_success": 0.0,
1154
+ "base_utility": 0.07191530615091324,
1155
+ "coverage": 0.0,
1156
+ "execute_generated": 0.0,
1157
+ "fallback_rate": 1.0,
1158
+ "hidden_chart_oracle_success": 1.0,
1159
+ "hidden_chart_oracle_utility": 2.0,
1160
+ "outcome_ptr": 1.0,
1161
+ "predicted_margin": 0.2393376995840159,
1162
+ "proposal_oracle_success": 1.0,
1163
+ "proposal_oracle_utility": 2.0,
1164
+ "selected_candidate_index": 3.0,
1165
+ "selected_success": 0.0,
1166
+ "selected_success_gain_over_base": 0.0,
1167
+ "selected_utility": 0.07191530615091324,
1168
+ "selected_utility_gain_over_base": 0.0,
1169
+ "selector_regret": 1.9280846938490868,
1170
+ "success_selector_gap": 1.0,
1171
+ "success_support_gap": 0.0,
1172
+ "support_gap": 0.0,
1173
+ "tau": 0.27486681252026457
1174
+ },
1175
+ "1845489013/train0": {
1176
+ "base_success": 1.0,
1177
+ "base_utility": 2.0,
1178
+ "coverage": 1.0,
1179
+ "execute_generated": 1.0,
1180
+ "fallback_rate": 0.0,
1181
+ "hidden_chart_oracle_success": 1.0,
1182
+ "hidden_chart_oracle_utility": 2.0,
1183
+ "outcome_ptr": 0.0,
1184
+ "predicted_margin": 0.29215703842542395,
1185
+ "proposal_oracle_success": 1.0,
1186
+ "proposal_oracle_utility": 2.0,
1187
+ "selected_candidate_index": 1.0,
1188
+ "selected_success": 0.0,
1189
+ "selected_success_gain_over_base": -1.0,
1190
+ "selected_utility": 0.0578075647354126,
1191
+ "selected_utility_gain_over_base": -1.9421924352645874,
1192
+ "selector_regret": 1.9421924352645874,
1193
+ "success_selector_gap": 1.0,
1194
+ "success_support_gap": 0.0,
1195
+ "support_gap": 0.0,
1196
+ "tau": 0.27486681252026457
1197
+ },
1198
+ "1845489013/train1": {
1199
+ "base_success": 1.0,
1200
+ "base_utility": 2.0,
1201
+ "coverage": 1.0,
1202
+ "execute_generated": 1.0,
1203
+ "fallback_rate": 0.0,
1204
+ "hidden_chart_oracle_success": 1.0,
1205
+ "hidden_chart_oracle_utility": 2.0,
1206
+ "outcome_ptr": 0.0,
1207
+ "predicted_margin": 0.3205973255549167,
1208
+ "proposal_oracle_success": 1.0,
1209
+ "proposal_oracle_utility": 2.0,
1210
+ "selected_candidate_index": 1.0,
1211
+ "selected_success": 1.0,
1212
+ "selected_success_gain_over_base": 0.0,
1213
+ "selected_utility": 2.0,
1214
+ "selected_utility_gain_over_base": 0.0,
1215
+ "selector_regret": 0.0,
1216
+ "success_selector_gap": 0.0,
1217
+ "success_support_gap": 0.0,
1218
+ "support_gap": 0.0,
1219
+ "tau": 0.27486681252026457
1220
+ },
1221
+ "1845489013/train2": {
1222
+ "base_success": 1.0,
1223
+ "base_utility": 2.0,
1224
+ "coverage": 1.0,
1225
+ "execute_generated": 1.0,
1226
+ "fallback_rate": 0.0,
1227
+ "hidden_chart_oracle_success": 1.0,
1228
+ "hidden_chart_oracle_utility": 2.0,
1229
+ "outcome_ptr": 0.0,
1230
+ "predicted_margin": 0.3063515230930759,
1231
+ "proposal_oracle_success": 1.0,
1232
+ "proposal_oracle_utility": 2.0,
1233
+ "selected_candidate_index": 3.0,
1234
+ "selected_success": 1.0,
1235
+ "selected_success_gain_over_base": 0.0,
1236
+ "selected_utility": 2.0,
1237
+ "selected_utility_gain_over_base": 0.0,
1238
+ "selector_regret": 0.0,
1239
+ "success_selector_gap": 0.0,
1240
+ "success_support_gap": 0.0,
1241
+ "support_gap": 0.0,
1242
+ "tau": 0.27486681252026457
1243
+ },
1244
+ "1999022574/train0": {
1245
+ "base_success": 1.0,
1246
+ "base_utility": 2.0,
1247
+ "coverage": 1.0,
1248
+ "execute_generated": 1.0,
1249
+ "fallback_rate": 0.0,
1250
+ "hidden_chart_oracle_success": 1.0,
1251
+ "hidden_chart_oracle_utility": 2.0,
1252
+ "outcome_ptr": 0.0,
1253
+ "predicted_margin": 0.3315547259998632,
1254
+ "proposal_oracle_success": 1.0,
1255
+ "proposal_oracle_utility": 2.0,
1256
+ "selected_candidate_index": 4.0,
1257
+ "selected_success": 1.0,
1258
+ "selected_success_gain_over_base": 0.0,
1259
+ "selected_utility": 2.0,
1260
+ "selected_utility_gain_over_base": 0.0,
1261
+ "selector_regret": 0.0,
1262
+ "success_selector_gap": 0.0,
1263
+ "success_support_gap": 0.0,
1264
+ "support_gap": 0.0,
1265
+ "tau": 0.27486681252026457
1266
+ },
1267
+ "1999022574/train1": {
1268
+ "base_success": 1.0,
1269
+ "base_utility": 2.0,
1270
+ "coverage": 0.0,
1271
+ "execute_generated": 0.0,
1272
+ "fallback_rate": 1.0,
1273
+ "hidden_chart_oracle_success": 1.0,
1274
+ "hidden_chart_oracle_utility": 2.0,
1275
+ "outcome_ptr": 0.0,
1276
+ "predicted_margin": 0.26629497064359653,
1277
+ "proposal_oracle_success": 1.0,
1278
+ "proposal_oracle_utility": 2.0,
1279
+ "selected_candidate_index": 2.0,
1280
+ "selected_success": 1.0,
1281
+ "selected_success_gain_over_base": 0.0,
1282
+ "selected_utility": 2.0,
1283
+ "selected_utility_gain_over_base": 0.0,
1284
+ "selector_regret": 0.0,
1285
+ "success_selector_gap": 0.0,
1286
+ "success_support_gap": 0.0,
1287
+ "support_gap": 0.0,
1288
+ "tau": 0.27486681252026457
1289
+ },
1290
+ "1999022574/train2": {
1291
+ "base_success": 1.0,
1292
+ "base_utility": 2.0,
1293
+ "coverage": 1.0,
1294
+ "execute_generated": 1.0,
1295
+ "fallback_rate": 0.0,
1296
+ "hidden_chart_oracle_success": 1.0,
1297
+ "hidden_chart_oracle_utility": 2.0,
1298
+ "outcome_ptr": 0.0,
1299
+ "predicted_margin": 0.2860205906334836,
1300
+ "proposal_oracle_success": 1.0,
1301
+ "proposal_oracle_utility": 2.0,
1302
+ "selected_candidate_index": 2.0,
1303
+ "selected_success": 0.0,
1304
+ "selected_success_gain_over_base": -1.0,
1305
+ "selected_utility": 0.17059293389320374,
1306
+ "selected_utility_gain_over_base": -1.8294070661067963,
1307
+ "selector_regret": 1.8294070661067963,
1308
+ "success_selector_gap": 1.0,
1309
+ "success_support_gap": 0.0,
1310
+ "support_gap": 0.0,
1311
+ "tau": 0.27486681252026457
1312
+ },
1313
+ "2012211287/train0": {
1314
+ "base_success": 0.0,
1315
+ "base_utility": 0.7236053347587585,
1316
+ "coverage": 0.0,
1317
+ "execute_generated": 0.0,
1318
+ "fallback_rate": 1.0,
1319
+ "hidden_chart_oracle_success": 1.0,
1320
+ "hidden_chart_oracle_utility": 2.0,
1321
+ "outcome_ptr": 1.0,
1322
+ "predicted_margin": 0.2439288542794308,
1323
+ "proposal_oracle_success": 1.0,
1324
+ "proposal_oracle_utility": 2.0,
1325
+ "selected_candidate_index": 2.0,
1326
+ "selected_success": 0.0,
1327
+ "selected_success_gain_over_base": 0.0,
1328
+ "selected_utility": 0.7236053347587585,
1329
+ "selected_utility_gain_over_base": 0.0,
1330
+ "selector_regret": 1.2763946652412415,
1331
+ "success_selector_gap": 1.0,
1332
+ "success_support_gap": 0.0,
1333
+ "support_gap": 0.0,
1334
+ "tau": 0.27486681252026457
1335
+ },
1336
+ "2012211287/train1": {
1337
+ "base_success": 0.0,
1338
+ "base_utility": 0.7236053347587585,
1339
+ "coverage": 0.0,
1340
+ "execute_generated": 0.0,
1341
+ "fallback_rate": 1.0,
1342
+ "hidden_chart_oracle_success": 1.0,
1343
+ "hidden_chart_oracle_utility": 2.0,
1344
+ "outcome_ptr": 1.0,
1345
+ "predicted_margin": 0.2173899018320687,
1346
+ "proposal_oracle_success": 1.0,
1347
+ "proposal_oracle_utility": 2.0,
1348
+ "selected_candidate_index": 1.0,
1349
+ "selected_success": 0.0,
1350
+ "selected_success_gain_over_base": 0.0,
1351
+ "selected_utility": 0.7236053347587585,
1352
+ "selected_utility_gain_over_base": 0.0,
1353
+ "selector_regret": 1.2763946652412415,
1354
+ "success_selector_gap": 1.0,
1355
+ "success_support_gap": 0.0,
1356
+ "support_gap": 0.0,
1357
+ "tau": 0.27486681252026457
1358
+ },
1359
+ "2012211287/train2": {
1360
+ "base_success": 0.0,
1361
+ "base_utility": 0.7236053347587585,
1362
+ "coverage": 1.0,
1363
+ "execute_generated": 1.0,
1364
+ "fallback_rate": 0.0,
1365
+ "hidden_chart_oracle_success": 1.0,
1366
+ "hidden_chart_oracle_utility": 2.0,
1367
+ "outcome_ptr": 1.0,
1368
+ "predicted_margin": 0.2954961785954242,
1369
+ "proposal_oracle_success": 1.0,
1370
+ "proposal_oracle_utility": 2.0,
1371
+ "selected_candidate_index": 4.0,
1372
+ "selected_success": 1.0,
1373
+ "selected_success_gain_over_base": 1.0,
1374
+ "selected_utility": 2.0,
1375
+ "selected_utility_gain_over_base": 1.2763946652412415,
1376
+ "selector_regret": 0.0,
1377
+ "success_selector_gap": 0.0,
1378
+ "success_support_gap": 0.0,
1379
+ "support_gap": 0.0,
1380
+ "tau": 0.27486681252026457
1381
+ },
1382
+ "2038594191/train0": {
1383
+ "base_success": 1.0,
1384
+ "base_utility": 2.0,
1385
+ "coverage": 1.0,
1386
+ "execute_generated": 1.0,
1387
+ "fallback_rate": 0.0,
1388
+ "hidden_chart_oracle_success": 1.0,
1389
+ "hidden_chart_oracle_utility": 2.0,
1390
+ "outcome_ptr": 0.0,
1391
+ "predicted_margin": 0.2993318508024962,
1392
+ "proposal_oracle_success": 1.0,
1393
+ "proposal_oracle_utility": 2.0,
1394
+ "selected_candidate_index": 4.0,
1395
+ "selected_success": 1.0,
1396
+ "selected_success_gain_over_base": 0.0,
1397
+ "selected_utility": 2.0,
1398
+ "selected_utility_gain_over_base": 0.0,
1399
+ "selector_regret": 0.0,
1400
+ "success_selector_gap": 0.0,
1401
+ "success_support_gap": 0.0,
1402
+ "support_gap": 0.0,
1403
+ "tau": 0.27486681252026457
1404
+ },
1405
+ "2038594191/train1": {
1406
+ "base_success": 1.0,
1407
+ "base_utility": 2.0,
1408
+ "coverage": 1.0,
1409
+ "execute_generated": 1.0,
1410
+ "fallback_rate": 0.0,
1411
+ "hidden_chart_oracle_success": 1.0,
1412
+ "hidden_chart_oracle_utility": 2.0,
1413
+ "outcome_ptr": 0.0,
1414
+ "predicted_margin": 0.2979604464108771,
1415
+ "proposal_oracle_success": 1.0,
1416
+ "proposal_oracle_utility": 2.0,
1417
+ "selected_candidate_index": 5.0,
1418
+ "selected_success": 0.0,
1419
+ "selected_success_gain_over_base": -1.0,
1420
+ "selected_utility": 0.07674666494131088,
1421
+ "selected_utility_gain_over_base": -1.9232533350586891,
1422
+ "selector_regret": 1.9232533350586891,
1423
+ "success_selector_gap": 1.0,
1424
+ "success_support_gap": 0.0,
1425
+ "support_gap": 0.0,
1426
+ "tau": 0.27486681252026457
1427
+ },
1428
+ "2038594191/train2": {
1429
+ "base_success": 1.0,
1430
+ "base_utility": 2.0,
1431
+ "coverage": 1.0,
1432
+ "execute_generated": 1.0,
1433
+ "fallback_rate": 0.0,
1434
+ "hidden_chart_oracle_success": 1.0,
1435
+ "hidden_chart_oracle_utility": 2.0,
1436
+ "outcome_ptr": 0.0,
1437
+ "predicted_margin": 0.31805097808305605,
1438
+ "proposal_oracle_success": 1.0,
1439
+ "proposal_oracle_utility": 2.0,
1440
+ "selected_candidate_index": 1.0,
1441
+ "selected_success": 1.0,
1442
+ "selected_success_gain_over_base": 0.0,
1443
+ "selected_utility": 2.0,
1444
+ "selected_utility_gain_over_base": 0.0,
1445
+ "selector_regret": 0.0,
1446
+ "success_selector_gap": 0.0,
1447
+ "success_support_gap": 0.0,
1448
+ "support_gap": 0.0,
1449
+ "tau": 0.27486681252026457
1450
+ },
1451
+ "2054899126/train0": {
1452
+ "base_success": 1.0,
1453
+ "base_utility": 2.0,
1454
+ "coverage": 0.5,
1455
+ "execute_generated": 0.5,
1456
+ "fallback_rate": 0.5,
1457
+ "hidden_chart_oracle_success": 1.0,
1458
+ "hidden_chart_oracle_utility": 2.0,
1459
+ "outcome_ptr": 0.0,
1460
+ "predicted_margin": 0.29507378977597865,
1461
+ "proposal_oracle_success": 0.5,
1462
+ "proposal_oracle_utility": 1.077283151447773,
1463
+ "selected_candidate_index": 4.5,
1464
+ "selected_success": 1.0,
1465
+ "selected_success_gain_over_base": 0.0,
1466
+ "selected_utility": 2.0,
1467
+ "selected_utility_gain_over_base": 0.0,
1468
+ "selector_regret": 0.0,
1469
+ "success_selector_gap": 0.0,
1470
+ "success_support_gap": 0.5,
1471
+ "support_gap": 0.922716848552227,
1472
+ "tau": 0.27486681252026457
1473
+ },
1474
+ "2054899126/train1": {
1475
+ "base_success": 1.0,
1476
+ "base_utility": 2.0,
1477
+ "coverage": 0.5,
1478
+ "execute_generated": 0.5,
1479
+ "fallback_rate": 0.5,
1480
+ "hidden_chart_oracle_success": 1.0,
1481
+ "hidden_chart_oracle_utility": 2.0,
1482
+ "outcome_ptr": 0.0,
1483
+ "predicted_margin": 0.3147264310108414,
1484
+ "proposal_oracle_success": 0.5,
1485
+ "proposal_oracle_utility": 1.0696808770298958,
1486
+ "selected_candidate_index": 3.0,
1487
+ "selected_success": 1.0,
1488
+ "selected_success_gain_over_base": 0.0,
1489
+ "selected_utility": 2.0,
1490
+ "selected_utility_gain_over_base": 0.0,
1491
+ "selector_regret": 0.0,
1492
+ "success_selector_gap": 0.0,
1493
+ "success_support_gap": 0.5,
1494
+ "support_gap": 0.9303191229701042,
1495
+ "tau": 0.27486681252026457
1496
+ },
1497
+ "2054899126/train2": {
1498
+ "base_success": 1.0,
1499
+ "base_utility": 2.0,
1500
+ "coverage": 0.5,
1501
+ "execute_generated": 0.5,
1502
+ "fallback_rate": 0.5,
1503
+ "hidden_chart_oracle_success": 1.0,
1504
+ "hidden_chart_oracle_utility": 2.0,
1505
+ "outcome_ptr": 0.0,
1506
+ "predicted_margin": 0.26760955703093814,
1507
+ "proposal_oracle_success": 0.5,
1508
+ "proposal_oracle_utility": 1.0455262660980225,
1509
+ "selected_candidate_index": 0.0,
1510
+ "selected_success": 1.0,
1511
+ "selected_success_gain_over_base": 0.0,
1512
+ "selected_utility": 2.0,
1513
+ "selected_utility_gain_over_base": 0.0,
1514
+ "selector_regret": 0.0,
1515
+ "success_selector_gap": 0.0,
1516
+ "success_support_gap": 0.5,
1517
+ "support_gap": 0.9544737339019775,
1518
+ "tau": 0.27486681252026457
1519
+ },
1520
+ "2061635888/train0": {
1521
+ "base_success": 0.0,
1522
+ "base_utility": 0.05225050449371338,
1523
+ "coverage": 1.0,
1524
+ "execute_generated": 1.0,
1525
+ "fallback_rate": 0.0,
1526
+ "hidden_chart_oracle_success": 1.0,
1527
+ "hidden_chart_oracle_utility": 2.0,
1528
+ "outcome_ptr": 1.0,
1529
+ "predicted_margin": 0.2853496139550879,
1530
+ "proposal_oracle_success": 0.0,
1531
+ "proposal_oracle_utility": 0.09799930453300476,
1532
+ "selected_candidate_index": 2.0,
1533
+ "selected_success": 0.0,
1534
+ "selected_success_gain_over_base": 0.0,
1535
+ "selected_utility": 0.09799930453300476,
1536
+ "selected_utility_gain_over_base": 0.04574880003929138,
1537
+ "selector_regret": 0.0,
1538
+ "success_selector_gap": 0.0,
1539
+ "success_support_gap": 1.0,
1540
+ "support_gap": 1.9020006954669952,
1541
+ "tau": 0.27486681252026457
1542
+ },
1543
+ "2061635888/train1": {
1544
+ "base_success": 0.0,
1545
+ "base_utility": 0.05225050449371338,
1546
+ "coverage": 1.0,
1547
+ "execute_generated": 1.0,
1548
+ "fallback_rate": 0.0,
1549
+ "hidden_chart_oracle_success": 1.0,
1550
+ "hidden_chart_oracle_utility": 2.0,
1551
+ "outcome_ptr": 1.0,
1552
+ "predicted_margin": 0.30151450441950267,
1553
+ "proposal_oracle_success": 0.0,
1554
+ "proposal_oracle_utility": 0.07580336928367615,
1555
+ "selected_candidate_index": 2.0,
1556
+ "selected_success": 0.0,
1557
+ "selected_success_gain_over_base": 0.0,
1558
+ "selected_utility": 0.07580336928367615,
1559
+ "selected_utility_gain_over_base": 0.02355286478996277,
1560
+ "selector_regret": 0.0,
1561
+ "success_selector_gap": 0.0,
1562
+ "success_support_gap": 1.0,
1563
+ "support_gap": 1.9241966307163239,
1564
+ "tau": 0.27486681252026457
1565
+ },
1566
+ "2061635888/train2": {
1567
+ "base_success": 0.0,
1568
+ "base_utility": 0.05225050449371338,
1569
+ "coverage": 1.0,
1570
+ "execute_generated": 1.0,
1571
+ "fallback_rate": 0.0,
1572
+ "hidden_chart_oracle_success": 1.0,
1573
+ "hidden_chart_oracle_utility": 2.0,
1574
+ "outcome_ptr": 1.0,
1575
+ "predicted_margin": 0.2939770755284871,
1576
+ "proposal_oracle_success": 0.0,
1577
+ "proposal_oracle_utility": 0.11105132102966309,
1578
+ "selected_candidate_index": 4.0,
1579
+ "selected_success": 0.0,
1580
+ "selected_success_gain_over_base": 0.0,
1581
+ "selected_utility": 0.11105132102966309,
1582
+ "selected_utility_gain_over_base": 0.05880081653594971,
1583
+ "selector_regret": 0.0,
1584
+ "success_selector_gap": 0.0,
1585
+ "success_support_gap": 1.0,
1586
+ "support_gap": 1.888948678970337,
1587
+ "tau": 0.27486681252026457
1588
+ },
1589
+ "211245842/train0": {
1590
+ "base_success": 0.0,
1591
+ "base_utility": 0.19090841710567474,
1592
+ "coverage": 1.0,
1593
+ "execute_generated": 1.0,
1594
+ "fallback_rate": 0.0,
1595
+ "hidden_chart_oracle_success": 1.0,
1596
+ "hidden_chart_oracle_utility": 2.0,
1597
+ "outcome_ptr": 1.0,
1598
+ "predicted_margin": 0.3480361129591947,
1599
+ "proposal_oracle_success": 1.0,
1600
+ "proposal_oracle_utility": 2.0,
1601
+ "selected_candidate_index": 2.0,
1602
+ "selected_success": 0.0,
1603
+ "selected_success_gain_over_base": 0.0,
1604
+ "selected_utility": 0.0942162424325943,
1605
+ "selected_utility_gain_over_base": -0.09669217467308044,
1606
+ "selector_regret": 1.9057837575674057,
1607
+ "success_selector_gap": 1.0,
1608
+ "success_support_gap": 0.0,
1609
+ "support_gap": 0.0,
1610
+ "tau": 0.27486681252026457
1611
+ },
1612
+ "211245842/train1": {
1613
+ "base_success": 0.0,
1614
+ "base_utility": 0.19090841710567474,
1615
+ "coverage": 0.0,
1616
+ "execute_generated": 0.0,
1617
+ "fallback_rate": 1.0,
1618
+ "hidden_chart_oracle_success": 1.0,
1619
+ "hidden_chart_oracle_utility": 2.0,
1620
+ "outcome_ptr": 1.0,
1621
+ "predicted_margin": 0.2682947398568944,
1622
+ "proposal_oracle_success": 0.0,
1623
+ "proposal_oracle_utility": 0.21295717358589172,
1624
+ "selected_candidate_index": 2.0,
1625
+ "selected_success": 0.0,
1626
+ "selected_success_gain_over_base": 0.0,
1627
+ "selected_utility": 0.19090841710567474,
1628
+ "selected_utility_gain_over_base": 0.0,
1629
+ "selector_regret": 0.02204875648021698,
1630
+ "success_selector_gap": 0.0,
1631
+ "success_support_gap": 1.0,
1632
+ "support_gap": 1.7870428264141083,
1633
+ "tau": 0.27486681252026457
1634
+ },
1635
+ "211245842/train2": {
1636
+ "base_success": 0.0,
1637
+ "base_utility": 0.19090841710567474,
1638
+ "coverage": 1.0,
1639
+ "execute_generated": 1.0,
1640
+ "fallback_rate": 0.0,
1641
+ "hidden_chart_oracle_success": 1.0,
1642
+ "hidden_chart_oracle_utility": 2.0,
1643
+ "outcome_ptr": 1.0,
1644
+ "predicted_margin": 0.3359191058246588,
1645
+ "proposal_oracle_success": 0.0,
1646
+ "proposal_oracle_utility": 0.21523693203926086,
1647
+ "selected_candidate_index": 2.0,
1648
+ "selected_success": 0.0,
1649
+ "selected_success_gain_over_base": 0.0,
1650
+ "selected_utility": 0.21523693203926086,
1651
+ "selected_utility_gain_over_base": 0.02432851493358612,
1652
+ "selector_regret": 0.0,
1653
+ "success_selector_gap": 0.0,
1654
+ "success_support_gap": 1.0,
1655
+ "support_gap": 1.7847630679607391,
1656
+ "tau": 0.27486681252026457
1657
+ },
1658
+ "278008524/train0": {
1659
+ "base_success": 0.0,
1660
+ "base_utility": 0.042734989896416664,
1661
+ "coverage": 0.5,
1662
+ "execute_generated": 0.5,
1663
+ "fallback_rate": 0.5,
1664
+ "hidden_chart_oracle_success": 0.5,
1665
+ "hidden_chart_oracle_utility": 1.1139086410403252,
1666
+ "outcome_ptr": 1.0,
1667
+ "predicted_margin": 0.2678201102965153,
1668
+ "proposal_oracle_success": 0.5,
1669
+ "proposal_oracle_utility": 1.0287093725055456,
1670
+ "selected_candidate_index": 2.0,
1671
+ "selected_success": 0.0,
1672
+ "selected_success_gain_over_base": 0.0,
1673
+ "selected_utility": 0.04390883445739746,
1674
+ "selected_utility_gain_over_base": 0.0011738445609807968,
1675
+ "selector_regret": 0.9848005380481482,
1676
+ "success_selector_gap": 0.5,
1677
+ "success_support_gap": 0.0,
1678
+ "support_gap": 0.08519926853477955,
1679
+ "tau": 0.27486681252026457
1680
+ },
1681
+ "278008524/train1": {
1682
+ "base_success": 0.0,
1683
+ "base_utility": 0.042734989896416664,
1684
+ "coverage": 0.5,
1685
+ "execute_generated": 0.5,
1686
+ "fallback_rate": 0.5,
1687
+ "hidden_chart_oracle_success": 0.5,
1688
+ "hidden_chart_oracle_utility": 1.1139086410403252,
1689
+ "outcome_ptr": 1.0,
1690
+ "predicted_margin": 0.2877655933965288,
1691
+ "proposal_oracle_success": 0.5,
1692
+ "proposal_oracle_utility": 1.0376745946705341,
1693
+ "selected_candidate_index": 4.0,
1694
+ "selected_success": 0.0,
1695
+ "selected_success_gain_over_base": 0.0,
1696
+ "selected_utility": 0.05399738810956478,
1697
+ "selected_utility_gain_over_base": 0.011262398213148117,
1698
+ "selector_regret": 0.9836772065609694,
1699
+ "success_selector_gap": 0.5,
1700
+ "success_support_gap": 0.0,
1701
+ "support_gap": 0.07623404636979103,
1702
+ "tau": 0.27486681252026457
1703
+ },
1704
+ "278008524/train2": {
1705
+ "base_success": 0.0,
1706
+ "base_utility": 0.042734989896416664,
1707
+ "coverage": 0.5,
1708
+ "execute_generated": 0.5,
1709
+ "fallback_rate": 0.5,
1710
+ "hidden_chart_oracle_success": 0.5,
1711
+ "hidden_chart_oracle_utility": 1.1139086410403252,
1712
+ "outcome_ptr": 1.0,
1713
+ "predicted_margin": 0.27575955987422557,
1714
+ "proposal_oracle_success": 0.5,
1715
+ "proposal_oracle_utility": 1.1359489113092422,
1716
+ "selected_candidate_index": 3.0,
1717
+ "selected_success": 0.0,
1718
+ "selected_success_gain_over_base": 0.0,
1719
+ "selected_utility": 0.04342125728726387,
1720
+ "selected_utility_gain_over_base": 0.0006862673908472061,
1721
+ "selector_regret": 1.0925276540219784,
1722
+ "success_selector_gap": 0.5,
1723
+ "success_support_gap": 0.0,
1724
+ "support_gap": 0.0,
1725
+ "tau": 0.27486681252026457
1726
+ },
1727
+ "307068879/train0": {
1728
+ "base_success": 0.0,
1729
+ "base_utility": 0.6000905632972717,
1730
+ "coverage": 1.0,
1731
+ "execute_generated": 1.0,
1732
+ "fallback_rate": 0.0,
1733
+ "hidden_chart_oracle_success": 1.0,
1734
+ "hidden_chart_oracle_utility": 2.0,
1735
+ "outcome_ptr": 0.0,
1736
+ "predicted_margin": 0.27695436930521883,
1737
+ "proposal_oracle_success": 0.0,
1738
+ "proposal_oracle_utility": 0.2115575522184372,
1739
+ "selected_candidate_index": 2.0,
1740
+ "selected_success": 0.0,
1741
+ "selected_success_gain_over_base": 0.0,
1742
+ "selected_utility": 0.19644933938980103,
1743
+ "selected_utility_gain_over_base": -0.4036412239074707,
1744
+ "selector_regret": 0.01510821282863617,
1745
+ "success_selector_gap": 0.0,
1746
+ "success_support_gap": 1.0,
1747
+ "support_gap": 1.7884424477815628,
1748
+ "tau": 0.27486681252026457
1749
+ },
1750
+ "307068879/train1": {
1751
+ "base_success": 0.0,
1752
+ "base_utility": 0.6000905632972717,
1753
+ "coverage": 0.0,
1754
+ "execute_generated": 0.0,
1755
+ "fallback_rate": 1.0,
1756
+ "hidden_chart_oracle_success": 1.0,
1757
+ "hidden_chart_oracle_utility": 2.0,
1758
+ "outcome_ptr": 0.0,
1759
+ "predicted_margin": 0.24557416434177576,
1760
+ "proposal_oracle_success": 0.0,
1761
+ "proposal_oracle_utility": 0.5876073241233826,
1762
+ "selected_candidate_index": 1.0,
1763
+ "selected_success": 0.0,
1764
+ "selected_success_gain_over_base": 0.0,
1765
+ "selected_utility": 0.6000905632972717,
1766
+ "selected_utility_gain_over_base": 0.0,
1767
+ "selector_regret": 0.0,
1768
+ "success_selector_gap": 0.0,
1769
+ "success_support_gap": 1.0,
1770
+ "support_gap": 1.4123926758766174,
1771
+ "tau": 0.27486681252026457
1772
+ },
1773
+ "307068879/train2": {
1774
+ "base_success": 0.0,
1775
+ "base_utility": 0.6000905632972717,
1776
+ "coverage": 0.0,
1777
+ "execute_generated": 0.0,
1778
+ "fallback_rate": 1.0,
1779
+ "hidden_chart_oracle_success": 1.0,
1780
+ "hidden_chart_oracle_utility": 2.0,
1781
+ "outcome_ptr": 0.0,
1782
+ "predicted_margin": 0.2651483500538066,
1783
+ "proposal_oracle_success": 0.0,
1784
+ "proposal_oracle_utility": 0.22727864980697632,
1785
+ "selected_candidate_index": 6.0,
1786
+ "selected_success": 0.0,
1787
+ "selected_success_gain_over_base": 0.0,
1788
+ "selected_utility": 0.6000905632972717,
1789
+ "selected_utility_gain_over_base": 0.0,
1790
+ "selector_regret": 0.0,
1791
+ "success_selector_gap": 0.0,
1792
+ "success_support_gap": 1.0,
1793
+ "support_gap": 1.7727213501930237,
1794
+ "tau": 0.27486681252026457
1795
+ },
1796
+ "412102563/train0": {
1797
+ "base_success": 1.0,
1798
+ "base_utility": 2.0,
1799
+ "coverage": 1.0,
1800
+ "execute_generated": 1.0,
1801
+ "fallback_rate": 0.0,
1802
+ "hidden_chart_oracle_success": 1.0,
1803
+ "hidden_chart_oracle_utility": 2.0,
1804
+ "outcome_ptr": 0.0,
1805
+ "predicted_margin": 0.3414010961984837,
1806
+ "proposal_oracle_success": 1.0,
1807
+ "proposal_oracle_utility": 2.0,
1808
+ "selected_candidate_index": 3.0,
1809
+ "selected_success": 0.0,
1810
+ "selected_success_gain_over_base": -1.0,
1811
+ "selected_utility": 0.08495135605335236,
1812
+ "selected_utility_gain_over_base": -1.9150486439466476,
1813
+ "selector_regret": 1.9150486439466476,
1814
+ "success_selector_gap": 1.0,
1815
+ "success_support_gap": 0.0,
1816
+ "support_gap": 0.0,
1817
+ "tau": 0.27486681252026457
1818
+ },
1819
+ "412102563/train1": {
1820
+ "base_success": 1.0,
1821
+ "base_utility": 2.0,
1822
+ "coverage": 1.0,
1823
+ "execute_generated": 1.0,
1824
+ "fallback_rate": 0.0,
1825
+ "hidden_chart_oracle_success": 1.0,
1826
+ "hidden_chart_oracle_utility": 2.0,
1827
+ "outcome_ptr": 0.0,
1828
+ "predicted_margin": 0.2833111049570577,
1829
+ "proposal_oracle_success": 1.0,
1830
+ "proposal_oracle_utility": 2.0,
1831
+ "selected_candidate_index": 1.0,
1832
+ "selected_success": 1.0,
1833
+ "selected_success_gain_over_base": 0.0,
1834
+ "selected_utility": 2.0,
1835
+ "selected_utility_gain_over_base": 0.0,
1836
+ "selector_regret": 0.0,
1837
+ "success_selector_gap": 0.0,
1838
+ "success_support_gap": 0.0,
1839
+ "support_gap": 0.0,
1840
+ "tau": 0.27486681252026457
1841
+ },
1842
+ "412102563/train2": {
1843
+ "base_success": 1.0,
1844
+ "base_utility": 2.0,
1845
+ "coverage": 1.0,
1846
+ "execute_generated": 1.0,
1847
+ "fallback_rate": 0.0,
1848
+ "hidden_chart_oracle_success": 1.0,
1849
+ "hidden_chart_oracle_utility": 2.0,
1850
+ "outcome_ptr": 0.0,
1851
+ "predicted_margin": 0.28211686275122405,
1852
+ "proposal_oracle_success": 1.0,
1853
+ "proposal_oracle_utility": 2.0,
1854
+ "selected_candidate_index": 3.0,
1855
+ "selected_success": 0.0,
1856
+ "selected_success_gain_over_base": -1.0,
1857
+ "selected_utility": 0.812533438205719,
1858
+ "selected_utility_gain_over_base": -1.187466561794281,
1859
+ "selector_regret": 1.187466561794281,
1860
+ "success_selector_gap": 1.0,
1861
+ "success_support_gap": 0.0,
1862
+ "support_gap": 0.0,
1863
+ "tau": 0.27486681252026457
1864
+ },
1865
+ "440577711/train0": {
1866
+ "base_success": 0.0,
1867
+ "base_utility": 0.036152780055999756,
1868
+ "coverage": 1.0,
1869
+ "execute_generated": 1.0,
1870
+ "fallback_rate": 0.0,
1871
+ "hidden_chart_oracle_success": 1.0,
1872
+ "hidden_chart_oracle_utility": 2.0,
1873
+ "outcome_ptr": 1.0,
1874
+ "predicted_margin": 0.5373607975910129,
1875
+ "proposal_oracle_success": 1.0,
1876
+ "proposal_oracle_utility": 2.0,
1877
+ "selected_candidate_index": 1.0,
1878
+ "selected_success": 1.0,
1879
+ "selected_success_gain_over_base": 1.0,
1880
+ "selected_utility": 2.0,
1881
+ "selected_utility_gain_over_base": 1.9638472199440002,
1882
+ "selector_regret": 0.0,
1883
+ "success_selector_gap": 0.0,
1884
+ "success_support_gap": 0.0,
1885
+ "support_gap": 0.0,
1886
+ "tau": 0.27486681252026457
1887
+ },
1888
+ "440577711/train1": {
1889
+ "base_success": 0.0,
1890
+ "base_utility": 0.036152780055999756,
1891
+ "coverage": 1.0,
1892
+ "execute_generated": 1.0,
1893
+ "fallback_rate": 0.0,
1894
+ "hidden_chart_oracle_success": 1.0,
1895
+ "hidden_chart_oracle_utility": 2.0,
1896
+ "outcome_ptr": 1.0,
1897
+ "predicted_margin": 0.5500727644790677,
1898
+ "proposal_oracle_success": 1.0,
1899
+ "proposal_oracle_utility": 2.0,
1900
+ "selected_candidate_index": 2.0,
1901
+ "selected_success": 1.0,
1902
+ "selected_success_gain_over_base": 1.0,
1903
+ "selected_utility": 2.0,
1904
+ "selected_utility_gain_over_base": 1.9638472199440002,
1905
+ "selector_regret": 0.0,
1906
+ "success_selector_gap": 0.0,
1907
+ "success_support_gap": 0.0,
1908
+ "support_gap": 0.0,
1909
+ "tau": 0.27486681252026457
1910
+ },
1911
+ "440577711/train2": {
1912
+ "base_success": 0.0,
1913
+ "base_utility": 0.036152780055999756,
1914
+ "coverage": 1.0,
1915
+ "execute_generated": 1.0,
1916
+ "fallback_rate": 0.0,
1917
+ "hidden_chart_oracle_success": 1.0,
1918
+ "hidden_chart_oracle_utility": 2.0,
1919
+ "outcome_ptr": 1.0,
1920
+ "predicted_margin": 0.5748755452169654,
1921
+ "proposal_oracle_success": 1.0,
1922
+ "proposal_oracle_utility": 2.0,
1923
+ "selected_candidate_index": 1.0,
1924
+ "selected_success": 1.0,
1925
+ "selected_success_gain_over_base": 1.0,
1926
+ "selected_utility": 2.0,
1927
+ "selected_utility_gain_over_base": 1.9638472199440002,
1928
+ "selector_regret": 0.0,
1929
+ "success_selector_gap": 0.0,
1930
+ "success_support_gap": 0.0,
1931
+ "support_gap": 0.0,
1932
+ "tau": 0.27486681252026457
1933
+ },
1934
+ "441046872/train0": {
1935
+ "base_success": 0.0,
1936
+ "base_utility": 0.6999551653862,
1937
+ "coverage": 1.0,
1938
+ "execute_generated": 1.0,
1939
+ "fallback_rate": 0.0,
1940
+ "hidden_chart_oracle_success": 1.0,
1941
+ "hidden_chart_oracle_utility": 2.0,
1942
+ "outcome_ptr": 0.0,
1943
+ "predicted_margin": 0.29271657275824736,
1944
+ "proposal_oracle_success": 0.0,
1945
+ "proposal_oracle_utility": 0.5148025155067444,
1946
+ "selected_candidate_index": 2.0,
1947
+ "selected_success": 0.0,
1948
+ "selected_success_gain_over_base": 0.0,
1949
+ "selected_utility": 0.20389758050441742,
1950
+ "selected_utility_gain_over_base": -0.49605758488178253,
1951
+ "selector_regret": 0.31090493500232697,
1952
+ "success_selector_gap": 0.0,
1953
+ "success_support_gap": 1.0,
1954
+ "support_gap": 1.4851974844932556,
1955
+ "tau": 0.27486681252026457
1956
+ },
1957
+ "441046872/train1": {
1958
+ "base_success": 0.0,
1959
+ "base_utility": 0.6999551653862,
1960
+ "coverage": 0.0,
1961
+ "execute_generated": 0.0,
1962
+ "fallback_rate": 1.0,
1963
+ "hidden_chart_oracle_success": 1.0,
1964
+ "hidden_chart_oracle_utility": 2.0,
1965
+ "outcome_ptr": 1.0,
1966
+ "predicted_margin": 0.22930118783187353,
1967
+ "proposal_oracle_success": 0.0,
1968
+ "proposal_oracle_utility": 0.7233588695526123,
1969
+ "selected_candidate_index": 7.0,
1970
+ "selected_success": 0.0,
1971
+ "selected_success_gain_over_base": 0.0,
1972
+ "selected_utility": 0.6999551653862,
1973
+ "selected_utility_gain_over_base": 0.0,
1974
+ "selector_regret": 0.023403704166412354,
1975
+ "success_selector_gap": 0.0,
1976
+ "success_support_gap": 1.0,
1977
+ "support_gap": 1.2766411304473877,
1978
+ "tau": 0.27486681252026457
1979
+ },
1980
+ "441046872/train2": {
1981
+ "base_success": 0.0,
1982
+ "base_utility": 0.6999551653862,
1983
+ "coverage": 0.0,
1984
+ "execute_generated": 0.0,
1985
+ "fallback_rate": 1.0,
1986
+ "hidden_chart_oracle_success": 1.0,
1987
+ "hidden_chart_oracle_utility": 2.0,
1988
+ "outcome_ptr": 1.0,
1989
+ "predicted_margin": 0.2652207836698539,
1990
+ "proposal_oracle_success": 0.0,
1991
+ "proposal_oracle_utility": 0.7130289077758789,
1992
+ "selected_candidate_index": 5.0,
1993
+ "selected_success": 0.0,
1994
+ "selected_success_gain_over_base": 0.0,
1995
+ "selected_utility": 0.6999551653862,
1996
+ "selected_utility_gain_over_base": 0.0,
1997
+ "selector_regret": 0.013073742389678955,
1998
+ "success_selector_gap": 0.0,
1999
+ "success_support_gap": 1.0,
2000
+ "support_gap": 1.286971092224121,
2001
+ "tau": 0.27486681252026457
2002
+ },
2003
+ "509430392/train0": {
2004
+ "base_success": 0.0,
2005
+ "base_utility": 2.2351741790771484e-07,
2006
+ "coverage": 1.0,
2007
+ "execute_generated": 1.0,
2008
+ "fallback_rate": 0.0,
2009
+ "hidden_chart_oracle_success": 1.0,
2010
+ "hidden_chart_oracle_utility": 2.0,
2011
+ "outcome_ptr": 1.0,
2012
+ "predicted_margin": 0.48880504311757667,
2013
+ "proposal_oracle_success": 1.0,
2014
+ "proposal_oracle_utility": 2.0,
2015
+ "selected_candidate_index": 2.0,
2016
+ "selected_success": 0.0,
2017
+ "selected_success_gain_over_base": 0.0,
2018
+ "selected_utility": 6.729364395141602e-05,
2019
+ "selected_utility_gain_over_base": 6.70701265335083e-05,
2020
+ "selector_regret": 1.9999327063560486,
2021
+ "success_selector_gap": 1.0,
2022
+ "success_support_gap": 0.0,
2023
+ "support_gap": 0.0,
2024
+ "tau": 0.27486681252026457
2025
+ },
2026
+ "509430392/train1": {
2027
+ "base_success": 0.0,
2028
+ "base_utility": 2.2351741790771484e-07,
2029
+ "coverage": 1.0,
2030
+ "execute_generated": 1.0,
2031
+ "fallback_rate": 0.0,
2032
+ "hidden_chart_oracle_success": 1.0,
2033
+ "hidden_chart_oracle_utility": 2.0,
2034
+ "outcome_ptr": 1.0,
2035
+ "predicted_margin": 0.4752155661626043,
2036
+ "proposal_oracle_success": 0.0,
2037
+ "proposal_oracle_utility": 6.164610385894775e-05,
2038
+ "selected_candidate_index": 2.0,
2039
+ "selected_success": 0.0,
2040
+ "selected_success_gain_over_base": 0.0,
2041
+ "selected_utility": 1.519918441772461e-06,
2042
+ "selected_utility_gain_over_base": 1.296401023864746e-06,
2043
+ "selector_regret": 6.012618541717529e-05,
2044
+ "success_selector_gap": 0.0,
2045
+ "success_support_gap": 1.0,
2046
+ "support_gap": 1.999938353896141,
2047
+ "tau": 0.27486681252026457
2048
+ },
2049
+ "509430392/train2": {
2050
+ "base_success": 0.0,
2051
+ "base_utility": 2.2351741790771484e-07,
2052
+ "coverage": 1.0,
2053
+ "execute_generated": 1.0,
2054
+ "fallback_rate": 0.0,
2055
+ "hidden_chart_oracle_success": 1.0,
2056
+ "hidden_chart_oracle_utility": 2.0,
2057
+ "outcome_ptr": 1.0,
2058
+ "predicted_margin": 0.4695870241650073,
2059
+ "proposal_oracle_success": 0.0,
2060
+ "proposal_oracle_utility": 5.424022674560547e-06,
2061
+ "selected_candidate_index": 2.0,
2062
+ "selected_success": 0.0,
2063
+ "selected_success_gain_over_base": 0.0,
2064
+ "selected_utility": 1.4007091522216797e-06,
2065
+ "selected_utility_gain_over_base": 1.1771917343139648e-06,
2066
+ "selector_regret": 4.023313522338867e-06,
2067
+ "success_selector_gap": 0.0,
2068
+ "success_support_gap": 1.0,
2069
+ "support_gap": 1.9999945759773254,
2070
+ "tau": 0.27486681252026457
2071
+ },
2072
+ "585904130/train0": {
2073
+ "base_success": 0.0,
2074
+ "base_utility": 0.18171554803848267,
2075
+ "coverage": 0.0,
2076
+ "execute_generated": 0.0,
2077
+ "fallback_rate": 1.0,
2078
+ "hidden_chart_oracle_success": 1.0,
2079
+ "hidden_chart_oracle_utility": 2.0,
2080
+ "outcome_ptr": 1.0,
2081
+ "predicted_margin": 0.2588684275369812,
2082
+ "proposal_oracle_success": 1.0,
2083
+ "proposal_oracle_utility": 2.0,
2084
+ "selected_candidate_index": 7.0,
2085
+ "selected_success": 0.0,
2086
+ "selected_success_gain_over_base": 0.0,
2087
+ "selected_utility": 0.18171554803848267,
2088
+ "selected_utility_gain_over_base": 0.0,
2089
+ "selector_regret": 1.8182844519615173,
2090
+ "success_selector_gap": 1.0,
2091
+ "success_support_gap": 0.0,
2092
+ "support_gap": 0.0,
2093
+ "tau": 0.27486681252026457
2094
+ },
2095
+ "585904130/train1": {
2096
+ "base_success": 0.0,
2097
+ "base_utility": 0.18171554803848267,
2098
+ "coverage": 0.0,
2099
+ "execute_generated": 0.0,
2100
+ "fallback_rate": 1.0,
2101
+ "hidden_chart_oracle_success": 1.0,
2102
+ "hidden_chart_oracle_utility": 2.0,
2103
+ "outcome_ptr": 1.0,
2104
+ "predicted_margin": 0.25973841367102646,
2105
+ "proposal_oracle_success": 1.0,
2106
+ "proposal_oracle_utility": 2.0,
2107
+ "selected_candidate_index": 4.0,
2108
+ "selected_success": 0.0,
2109
+ "selected_success_gain_over_base": 0.0,
2110
+ "selected_utility": 0.18171554803848267,
2111
+ "selected_utility_gain_over_base": 0.0,
2112
+ "selector_regret": 1.8182844519615173,
2113
+ "success_selector_gap": 1.0,
2114
+ "success_support_gap": 0.0,
2115
+ "support_gap": 0.0,
2116
+ "tau": 0.27486681252026457
2117
+ },
2118
+ "585904130/train2": {
2119
+ "base_success": 0.0,
2120
+ "base_utility": 0.18171554803848267,
2121
+ "coverage": 1.0,
2122
+ "execute_generated": 1.0,
2123
+ "fallback_rate": 0.0,
2124
+ "hidden_chart_oracle_success": 1.0,
2125
+ "hidden_chart_oracle_utility": 2.0,
2126
+ "outcome_ptr": 1.0,
2127
+ "predicted_margin": 0.2959953103591755,
2128
+ "proposal_oracle_success": 1.0,
2129
+ "proposal_oracle_utility": 2.0,
2130
+ "selected_candidate_index": 1.0,
2131
+ "selected_success": 0.0,
2132
+ "selected_success_gain_over_base": 0.0,
2133
+ "selected_utility": 0.08694672584533691,
2134
+ "selected_utility_gain_over_base": -0.09476882219314575,
2135
+ "selector_regret": 1.913053274154663,
2136
+ "success_selector_gap": 1.0,
2137
+ "success_support_gap": 0.0,
2138
+ "support_gap": 0.0,
2139
+ "tau": 0.27486681252026457
2140
+ },
2141
+ "608016290/train0": {
2142
+ "base_success": 0.0,
2143
+ "base_utility": 0.725701630115509,
2144
+ "coverage": 1.0,
2145
+ "execute_generated": 1.0,
2146
+ "fallback_rate": 0.0,
2147
+ "hidden_chart_oracle_success": 1.0,
2148
+ "hidden_chart_oracle_utility": 2.0,
2149
+ "outcome_ptr": 1.0,
2150
+ "predicted_margin": 0.29600744925888234,
2151
+ "proposal_oracle_success": 1.0,
2152
+ "proposal_oracle_utility": 2.0,
2153
+ "selected_candidate_index": 7.0,
2154
+ "selected_success": 0.0,
2155
+ "selected_success_gain_over_base": 0.0,
2156
+ "selected_utility": 0.21135586500167847,
2157
+ "selected_utility_gain_over_base": -0.5143457651138306,
2158
+ "selector_regret": 1.7886441349983215,
2159
+ "success_selector_gap": 1.0,
2160
+ "success_support_gap": 0.0,
2161
+ "support_gap": 0.0,
2162
+ "tau": 0.27486681252026457
2163
+ },
2164
+ "608016290/train1": {
2165
+ "base_success": 0.0,
2166
+ "base_utility": 0.725701630115509,
2167
+ "coverage": 0.0,
2168
+ "execute_generated": 0.0,
2169
+ "fallback_rate": 1.0,
2170
+ "hidden_chart_oracle_success": 1.0,
2171
+ "hidden_chart_oracle_utility": 2.0,
2172
+ "outcome_ptr": 1.0,
2173
+ "predicted_margin": 0.2724535502374203,
2174
+ "proposal_oracle_success": 1.0,
2175
+ "proposal_oracle_utility": 2.0,
2176
+ "selected_candidate_index": 1.0,
2177
+ "selected_success": 0.0,
2178
+ "selected_success_gain_over_base": 0.0,
2179
+ "selected_utility": 0.725701630115509,
2180
+ "selected_utility_gain_over_base": 0.0,
2181
+ "selector_regret": 1.274298369884491,
2182
+ "success_selector_gap": 1.0,
2183
+ "success_support_gap": 0.0,
2184
+ "support_gap": 0.0,
2185
+ "tau": 0.27486681252026457
2186
+ },
2187
+ "608016290/train2": {
2188
+ "base_success": 0.0,
2189
+ "base_utility": 0.725701630115509,
2190
+ "coverage": 0.0,
2191
+ "execute_generated": 0.0,
2192
+ "fallback_rate": 1.0,
2193
+ "hidden_chart_oracle_success": 1.0,
2194
+ "hidden_chart_oracle_utility": 2.0,
2195
+ "outcome_ptr": 0.0,
2196
+ "predicted_margin": 0.2690682775496972,
2197
+ "proposal_oracle_success": 0.0,
2198
+ "proposal_oracle_utility": 0.6825780272483826,
2199
+ "selected_candidate_index": 2.0,
2200
+ "selected_success": 0.0,
2201
+ "selected_success_gain_over_base": 0.0,
2202
+ "selected_utility": 0.725701630115509,
2203
+ "selected_utility_gain_over_base": 0.0,
2204
+ "selector_regret": 0.0,
2205
+ "success_selector_gap": 0.0,
2206
+ "success_support_gap": 1.0,
2207
+ "support_gap": 1.3174219727516174,
2208
+ "tau": 0.27486681252026457
2209
+ },
2210
+ "626486813/train0": {
2211
+ "base_success": 0.0,
2212
+ "base_utility": 0.00509834272088483,
2213
+ "coverage": 0.0,
2214
+ "execute_generated": 0.0,
2215
+ "fallback_rate": 1.0,
2216
+ "hidden_chart_oracle_success": 0.0,
2217
+ "hidden_chart_oracle_utility": 0.16558238118886948,
2218
+ "outcome_ptr": 1.0,
2219
+ "predicted_margin": 0.2268756783853227,
2220
+ "proposal_oracle_success": 0.0,
2221
+ "proposal_oracle_utility": 0.043898362666368484,
2222
+ "selected_candidate_index": 4.0,
2223
+ "selected_success": 0.0,
2224
+ "selected_success_gain_over_base": 0.0,
2225
+ "selected_utility": 0.00509834272088483,
2226
+ "selected_utility_gain_over_base": 0.0,
2227
+ "selector_regret": 0.038800019945483655,
2228
+ "success_selector_gap": 0.0,
2229
+ "success_support_gap": 0.0,
2230
+ "support_gap": 0.12241578102111816,
2231
+ "tau": 0.27486681252026457
2232
+ },
2233
+ "626486813/train1": {
2234
+ "base_success": 0.0,
2235
+ "base_utility": 0.00509834272088483,
2236
+ "coverage": 0.0,
2237
+ "execute_generated": 0.0,
2238
+ "fallback_rate": 1.0,
2239
+ "hidden_chart_oracle_success": 0.0,
2240
+ "hidden_chart_oracle_utility": 0.16558238118886948,
2241
+ "outcome_ptr": 1.0,
2242
+ "predicted_margin": 0.2404699305536791,
2243
+ "proposal_oracle_success": 0.0,
2244
+ "proposal_oracle_utility": 0.030555277597159147,
2245
+ "selected_candidate_index": 4.0,
2246
+ "selected_success": 0.0,
2247
+ "selected_success_gain_over_base": 0.0,
2248
+ "selected_utility": 0.00509834272088483,
2249
+ "selected_utility_gain_over_base": 0.0,
2250
+ "selector_regret": 0.025456934876274318,
2251
+ "success_selector_gap": 0.0,
2252
+ "success_support_gap": 0.0,
2253
+ "support_gap": 0.13502710359171033,
2254
+ "tau": 0.27486681252026457
2255
+ },
2256
+ "626486813/train2": {
2257
+ "base_success": 0.0,
2258
+ "base_utility": 0.00509834272088483,
2259
+ "coverage": 1.0,
2260
+ "execute_generated": 1.0,
2261
+ "fallback_rate": 0.0,
2262
+ "hidden_chart_oracle_success": 0.0,
2263
+ "hidden_chart_oracle_utility": 0.16558238118886948,
2264
+ "outcome_ptr": 1.0,
2265
+ "predicted_margin": 0.29430194578042657,
2266
+ "proposal_oracle_success": 0.0,
2267
+ "proposal_oracle_utility": 0.03856666944921017,
2268
+ "selected_candidate_index": 2.0,
2269
+ "selected_success": 0.0,
2270
+ "selected_success_gain_over_base": 0.0,
2271
+ "selected_utility": 0.021789689548313618,
2272
+ "selected_utility_gain_over_base": 0.016691346827428788,
2273
+ "selector_regret": 0.01677697990089655,
2274
+ "success_selector_gap": 0.0,
2275
+ "success_support_gap": 0.0,
2276
+ "support_gap": 0.1270157117396593,
2277
+ "tau": 0.27486681252026457
2278
+ },
2279
+ "636680137/train0": {
2280
+ "base_success": 0.0,
2281
+ "base_utility": 0.08313504606485367,
2282
+ "coverage": 0.0,
2283
+ "execute_generated": 0.0,
2284
+ "fallback_rate": 1.0,
2285
+ "hidden_chart_oracle_success": 0.0,
2286
+ "hidden_chart_oracle_utility": 0.17056432366371155,
2287
+ "outcome_ptr": 1.0,
2288
+ "predicted_margin": 0.27089169668591645,
2289
+ "proposal_oracle_success": 0.0,
2290
+ "proposal_oracle_utility": 0.08826130628585815,
2291
+ "selected_candidate_index": 1.0,
2292
+ "selected_success": 0.0,
2293
+ "selected_success_gain_over_base": 0.0,
2294
+ "selected_utility": 0.08313504606485367,
2295
+ "selected_utility_gain_over_base": 0.0,
2296
+ "selector_regret": 0.005126260221004486,
2297
+ "success_selector_gap": 0.0,
2298
+ "success_support_gap": 0.0,
2299
+ "support_gap": 0.0823030173778534,
2300
+ "tau": 0.27486681252026457
2301
+ },
2302
+ "636680137/train1": {
2303
+ "base_success": 0.0,
2304
+ "base_utility": 0.08313504606485367,
2305
+ "coverage": 0.0,
2306
+ "execute_generated": 0.0,
2307
+ "fallback_rate": 1.0,
2308
+ "hidden_chart_oracle_success": 0.0,
2309
+ "hidden_chart_oracle_utility": 0.17056432366371155,
2310
+ "outcome_ptr": 0.0,
2311
+ "predicted_margin": 0.1973975413996576,
2312
+ "proposal_oracle_success": 0.0,
2313
+ "proposal_oracle_utility": 0.06741118431091309,
2314
+ "selected_candidate_index": 1.0,
2315
+ "selected_success": 0.0,
2316
+ "selected_success_gain_over_base": 0.0,
2317
+ "selected_utility": 0.08313504606485367,
2318
+ "selected_utility_gain_over_base": 0.0,
2319
+ "selector_regret": 0.0,
2320
+ "success_selector_gap": 0.0,
2321
+ "success_support_gap": 0.0,
2322
+ "support_gap": 0.10315313935279846,
2323
+ "tau": 0.27486681252026457
2324
+ },
2325
+ "636680137/train2": {
2326
+ "base_success": 0.0,
2327
+ "base_utility": 0.08313504606485367,
2328
+ "coverage": 0.0,
2329
+ "execute_generated": 0.0,
2330
+ "fallback_rate": 1.0,
2331
+ "hidden_chart_oracle_success": 0.0,
2332
+ "hidden_chart_oracle_utility": 0.17056432366371155,
2333
+ "outcome_ptr": 0.0,
2334
+ "predicted_margin": 0.25734913673209453,
2335
+ "proposal_oracle_success": 0.0,
2336
+ "proposal_oracle_utility": 0.0756516233086586,
2337
+ "selected_candidate_index": 7.0,
2338
+ "selected_success": 0.0,
2339
+ "selected_success_gain_over_base": 0.0,
2340
+ "selected_utility": 0.08313504606485367,
2341
+ "selected_utility_gain_over_base": 0.0,
2342
+ "selector_regret": 0.0,
2343
+ "success_selector_gap": 0.0,
2344
+ "success_support_gap": 0.0,
2345
+ "support_gap": 0.09491270035505295,
2346
+ "tau": 0.27486681252026457
2347
+ },
2348
+ "655884605/train0": {
2349
+ "base_success": 0.0,
2350
+ "base_utility": 0.16300667822360992,
2351
+ "coverage": 1.0,
2352
+ "execute_generated": 1.0,
2353
+ "fallback_rate": 0.0,
2354
+ "hidden_chart_oracle_success": 1.0,
2355
+ "hidden_chart_oracle_utility": 2.0,
2356
+ "outcome_ptr": 0.0,
2357
+ "predicted_margin": 0.29834663555269436,
2358
+ "proposal_oracle_success": 0.0,
2359
+ "proposal_oracle_utility": 0.13732868432998657,
2360
+ "selected_candidate_index": 2.0,
2361
+ "selected_success": 0.0,
2362
+ "selected_success_gain_over_base": 0.0,
2363
+ "selected_utility": 0.13732868432998657,
2364
+ "selected_utility_gain_over_base": -0.025677993893623352,
2365
+ "selector_regret": 0.0,
2366
+ "success_selector_gap": 0.0,
2367
+ "success_support_gap": 1.0,
2368
+ "support_gap": 1.8626713156700134,
2369
+ "tau": 0.27486681252026457
2370
+ },
2371
+ "655884605/train1": {
2372
+ "base_success": 0.0,
2373
+ "base_utility": 0.16300667822360992,
2374
+ "coverage": 1.0,
2375
+ "execute_generated": 1.0,
2376
+ "fallback_rate": 0.0,
2377
+ "hidden_chart_oracle_success": 1.0,
2378
+ "hidden_chart_oracle_utility": 2.0,
2379
+ "outcome_ptr": 1.0,
2380
+ "predicted_margin": 0.2923127432531102,
2381
+ "proposal_oracle_success": 0.0,
2382
+ "proposal_oracle_utility": 0.1651531159877777,
2383
+ "selected_candidate_index": 4.0,
2384
+ "selected_success": 0.0,
2385
+ "selected_success_gain_over_base": 0.0,
2386
+ "selected_utility": 0.1651531159877777,
2387
+ "selected_utility_gain_over_base": 0.0021464377641677856,
2388
+ "selector_regret": 0.0,
2389
+ "success_selector_gap": 0.0,
2390
+ "success_support_gap": 1.0,
2391
+ "support_gap": 1.8348468840122223,
2392
+ "tau": 0.27486681252026457
2393
+ },
2394
+ "655884605/train2": {
2395
+ "base_success": 0.0,
2396
+ "base_utility": 0.16300667822360992,
2397
+ "coverage": 1.0,
2398
+ "execute_generated": 1.0,
2399
+ "fallback_rate": 0.0,
2400
+ "hidden_chart_oracle_success": 1.0,
2401
+ "hidden_chart_oracle_utility": 2.0,
2402
+ "outcome_ptr": 1.0,
2403
+ "predicted_margin": 0.30274185948471083,
2404
+ "proposal_oracle_success": 0.0,
2405
+ "proposal_oracle_utility": 0.21447330713272095,
2406
+ "selected_candidate_index": 4.0,
2407
+ "selected_success": 0.0,
2408
+ "selected_success_gain_over_base": 0.0,
2409
+ "selected_utility": 0.05868344008922577,
2410
+ "selected_utility_gain_over_base": -0.10432323813438416,
2411
+ "selector_regret": 0.15578986704349518,
2412
+ "success_selector_gap": 0.0,
2413
+ "success_support_gap": 1.0,
2414
+ "support_gap": 1.785526692867279,
2415
+ "tau": 0.27486681252026457
2416
+ },
2417
+ "672328262/train0": {
2418
+ "base_success": 1.0,
2419
+ "base_utility": 2.0,
2420
+ "coverage": 0.0,
2421
+ "execute_generated": 0.0,
2422
+ "fallback_rate": 1.0,
2423
+ "hidden_chart_oracle_success": 1.0,
2424
+ "hidden_chart_oracle_utility": 2.0,
2425
+ "outcome_ptr": 0.0,
2426
+ "predicted_margin": 0.2633093366847031,
2427
+ "proposal_oracle_success": 1.0,
2428
+ "proposal_oracle_utility": 2.0,
2429
+ "selected_candidate_index": 5.0,
2430
+ "selected_success": 1.0,
2431
+ "selected_success_gain_over_base": 0.0,
2432
+ "selected_utility": 2.0,
2433
+ "selected_utility_gain_over_base": 0.0,
2434
+ "selector_regret": 0.0,
2435
+ "success_selector_gap": 0.0,
2436
+ "success_support_gap": 0.0,
2437
+ "support_gap": 0.0,
2438
+ "tau": 0.27486681252026457
2439
+ },
2440
+ "672328262/train1": {
2441
+ "base_success": 1.0,
2442
+ "base_utility": 2.0,
2443
+ "coverage": 1.0,
2444
+ "execute_generated": 1.0,
2445
+ "fallback_rate": 0.0,
2446
+ "hidden_chart_oracle_success": 1.0,
2447
+ "hidden_chart_oracle_utility": 2.0,
2448
+ "outcome_ptr": 0.0,
2449
+ "predicted_margin": 0.28392141565358076,
2450
+ "proposal_oracle_success": 1.0,
2451
+ "proposal_oracle_utility": 2.0,
2452
+ "selected_candidate_index": 5.0,
2453
+ "selected_success": 0.0,
2454
+ "selected_success_gain_over_base": -1.0,
2455
+ "selected_utility": 0.051487941294908524,
2456
+ "selected_utility_gain_over_base": -1.9485120587050915,
2457
+ "selector_regret": 1.9485120587050915,
2458
+ "success_selector_gap": 1.0,
2459
+ "success_support_gap": 0.0,
2460
+ "support_gap": 0.0,
2461
+ "tau": 0.27486681252026457
2462
+ },
2463
+ "672328262/train2": {
2464
+ "base_success": 1.0,
2465
+ "base_utility": 2.0,
2466
+ "coverage": 1.0,
2467
+ "execute_generated": 1.0,
2468
+ "fallback_rate": 0.0,
2469
+ "hidden_chart_oracle_success": 1.0,
2470
+ "hidden_chart_oracle_utility": 2.0,
2471
+ "outcome_ptr": 0.0,
2472
+ "predicted_margin": 0.30346720323430953,
2473
+ "proposal_oracle_success": 1.0,
2474
+ "proposal_oracle_utility": 2.0,
2475
+ "selected_candidate_index": 7.0,
2476
+ "selected_success": 0.0,
2477
+ "selected_success_gain_over_base": -1.0,
2478
+ "selected_utility": 0.06921929121017456,
2479
+ "selected_utility_gain_over_base": -1.9307807087898254,
2480
+ "selector_regret": 1.9307807087898254,
2481
+ "success_selector_gap": 1.0,
2482
+ "success_support_gap": 0.0,
2483
+ "support_gap": 0.0,
2484
+ "tau": 0.27486681252026457
2485
+ },
2486
+ "701667486/train0": {
2487
+ "base_success": 1.0,
2488
+ "base_utility": 2.0,
2489
+ "coverage": 1.0,
2490
+ "execute_generated": 1.0,
2491
+ "fallback_rate": 0.0,
2492
+ "hidden_chart_oracle_success": 1.0,
2493
+ "hidden_chart_oracle_utility": 2.0,
2494
+ "outcome_ptr": 0.0,
2495
+ "predicted_margin": 0.2977412732319387,
2496
+ "proposal_oracle_success": 1.0,
2497
+ "proposal_oracle_utility": 2.0,
2498
+ "selected_candidate_index": 2.0,
2499
+ "selected_success": 1.0,
2500
+ "selected_success_gain_over_base": 0.0,
2501
+ "selected_utility": 2.0,
2502
+ "selected_utility_gain_over_base": 0.0,
2503
+ "selector_regret": 0.0,
2504
+ "success_selector_gap": 0.0,
2505
+ "success_support_gap": 0.0,
2506
+ "support_gap": 0.0,
2507
+ "tau": 0.27486681252026457
2508
+ },
2509
+ "701667486/train1": {
2510
+ "base_success": 1.0,
2511
+ "base_utility": 2.0,
2512
+ "coverage": 1.0,
2513
+ "execute_generated": 1.0,
2514
+ "fallback_rate": 0.0,
2515
+ "hidden_chart_oracle_success": 1.0,
2516
+ "hidden_chart_oracle_utility": 2.0,
2517
+ "outcome_ptr": 0.0,
2518
+ "predicted_margin": 0.3259625098446074,
2519
+ "proposal_oracle_success": 1.0,
2520
+ "proposal_oracle_utility": 2.0,
2521
+ "selected_candidate_index": 4.0,
2522
+ "selected_success": 1.0,
2523
+ "selected_success_gain_over_base": 0.0,
2524
+ "selected_utility": 2.0,
2525
+ "selected_utility_gain_over_base": 0.0,
2526
+ "selector_regret": 0.0,
2527
+ "success_selector_gap": 0.0,
2528
+ "success_support_gap": 0.0,
2529
+ "support_gap": 0.0,
2530
+ "tau": 0.27486681252026457
2531
+ },
2532
+ "701667486/train2": {
2533
+ "base_success": 1.0,
2534
+ "base_utility": 2.0,
2535
+ "coverage": 1.0,
2536
+ "execute_generated": 1.0,
2537
+ "fallback_rate": 0.0,
2538
+ "hidden_chart_oracle_success": 1.0,
2539
+ "hidden_chart_oracle_utility": 2.0,
2540
+ "outcome_ptr": 0.0,
2541
+ "predicted_margin": 0.3129761167097917,
2542
+ "proposal_oracle_success": 1.0,
2543
+ "proposal_oracle_utility": 2.0,
2544
+ "selected_candidate_index": 7.0,
2545
+ "selected_success": 0.0,
2546
+ "selected_success_gain_over_base": -1.0,
2547
+ "selected_utility": 0.8125,
2548
+ "selected_utility_gain_over_base": -1.1875,
2549
+ "selector_regret": 1.1875,
2550
+ "success_selector_gap": 1.0,
2551
+ "success_support_gap": 0.0,
2552
+ "support_gap": 0.0,
2553
+ "tau": 0.27486681252026457
2554
+ },
2555
+ "747260437/train0": {
2556
+ "base_success": 1.0,
2557
+ "base_utility": 2.0,
2558
+ "coverage": 1.0,
2559
+ "execute_generated": 1.0,
2560
+ "fallback_rate": 0.0,
2561
+ "hidden_chart_oracle_success": 0.0,
2562
+ "hidden_chart_oracle_utility": 0.2786632478237152,
2563
+ "outcome_ptr": 0.0,
2564
+ "predicted_margin": 0.2908157340438423,
2565
+ "proposal_oracle_success": 0.0,
2566
+ "proposal_oracle_utility": 0.10839563608169556,
2567
+ "selected_candidate_index": 2.0,
2568
+ "selected_success": 0.0,
2569
+ "selected_success_gain_over_base": -1.0,
2570
+ "selected_utility": 0.07023122161626816,
2571
+ "selected_utility_gain_over_base": -1.9297687783837318,
2572
+ "selector_regret": 0.0381644144654274,
2573
+ "success_selector_gap": 0.0,
2574
+ "success_support_gap": 0.0,
2575
+ "support_gap": 0.17026761174201965,
2576
+ "tau": 0.27486681252026457
2577
+ },
2578
+ "747260437/train1": {
2579
+ "base_success": 1.0,
2580
+ "base_utility": 2.0,
2581
+ "coverage": 0.0,
2582
+ "execute_generated": 0.0,
2583
+ "fallback_rate": 1.0,
2584
+ "hidden_chart_oracle_success": 0.0,
2585
+ "hidden_chart_oracle_utility": 0.2786632478237152,
2586
+ "outcome_ptr": 0.0,
2587
+ "predicted_margin": 0.2561021548189116,
2588
+ "proposal_oracle_success": 0.0,
2589
+ "proposal_oracle_utility": 0.05267423391342163,
2590
+ "selected_candidate_index": 4.0,
2591
+ "selected_success": 1.0,
2592
+ "selected_success_gain_over_base": 0.0,
2593
+ "selected_utility": 2.0,
2594
+ "selected_utility_gain_over_base": 0.0,
2595
+ "selector_regret": 0.0,
2596
+ "success_selector_gap": 0.0,
2597
+ "success_support_gap": 0.0,
2598
+ "support_gap": 0.22598901391029358,
2599
+ "tau": 0.27486681252026457
2600
+ },
2601
+ "747260437/train2": {
2602
+ "base_success": 1.0,
2603
+ "base_utility": 2.0,
2604
+ "coverage": 0.0,
2605
+ "execute_generated": 0.0,
2606
+ "fallback_rate": 1.0,
2607
+ "hidden_chart_oracle_success": 0.0,
2608
+ "hidden_chart_oracle_utility": 0.2786632478237152,
2609
+ "outcome_ptr": 0.0,
2610
+ "predicted_margin": 0.254550993030934,
2611
+ "proposal_oracle_success": 0.0,
2612
+ "proposal_oracle_utility": 0.12510187923908234,
2613
+ "selected_candidate_index": 4.0,
2614
+ "selected_success": 1.0,
2615
+ "selected_success_gain_over_base": 0.0,
2616
+ "selected_utility": 2.0,
2617
+ "selected_utility_gain_over_base": 0.0,
2618
+ "selector_regret": 0.0,
2619
+ "success_selector_gap": 0.0,
2620
+ "success_support_gap": 0.0,
2621
+ "support_gap": 0.15356136858463287,
2622
+ "tau": 0.27486681252026457
2623
+ },
2624
+ "750448296/train0": {
2625
+ "base_success": 0.0,
2626
+ "base_utility": 0.11759314934412639,
2627
+ "coverage": 0.3333333333333333,
2628
+ "execute_generated": 0.3333333333333333,
2629
+ "fallback_rate": 0.6666666666666666,
2630
+ "hidden_chart_oracle_success": 0.6666666666666666,
2631
+ "hidden_chart_oracle_utility": 1.4079964955647786,
2632
+ "outcome_ptr": 0.6666666666666666,
2633
+ "predicted_margin": 0.2711584188696943,
2634
+ "proposal_oracle_success": 0.3333333333333333,
2635
+ "proposal_oracle_utility": 0.7526603279014429,
2636
+ "selected_candidate_index": 1.0,
2637
+ "selected_success": 0.0,
2638
+ "selected_success_gain_over_base": 0.0,
2639
+ "selected_utility": 0.1003921168545882,
2640
+ "selected_utility_gain_over_base": -0.017201032489538193,
2641
+ "selector_regret": 0.6522682110468546,
2642
+ "success_selector_gap": 0.3333333333333333,
2643
+ "success_support_gap": 0.3333333333333333,
2644
+ "support_gap": 0.6553361676633358,
2645
+ "tau": 0.27486681252026457
2646
+ },
2647
+ "750448296/train1": {
2648
+ "base_success": 0.0,
2649
+ "base_utility": 0.11759314934412639,
2650
+ "coverage": 0.6666666666666666,
2651
+ "execute_generated": 0.6666666666666666,
2652
+ "fallback_rate": 0.3333333333333333,
2653
+ "hidden_chart_oracle_success": 0.6666666666666666,
2654
+ "hidden_chart_oracle_utility": 1.4079964955647786,
2655
+ "outcome_ptr": 1.0,
2656
+ "predicted_margin": 0.3074482659083981,
2657
+ "proposal_oracle_success": 0.6666666666666666,
2658
+ "proposal_oracle_utility": 1.3705032741030057,
2659
+ "selected_candidate_index": 4.0,
2660
+ "selected_success": 0.0,
2661
+ "selected_success_gain_over_base": 0.0,
2662
+ "selected_utility": 0.07925570011138916,
2663
+ "selected_utility_gain_over_base": -0.03833744923273722,
2664
+ "selector_regret": 1.2912475739916165,
2665
+ "success_selector_gap": 0.6666666666666666,
2666
+ "success_support_gap": 0.0,
2667
+ "support_gap": 0.03749322146177292,
2668
+ "tau": 0.27486681252026457
2669
+ },
2670
+ "750448296/train2": {
2671
+ "base_success": 0.0,
2672
+ "base_utility": 0.11759314934412639,
2673
+ "coverage": 0.6666666666666666,
2674
+ "execute_generated": 0.6666666666666666,
2675
+ "fallback_rate": 0.3333333333333333,
2676
+ "hidden_chart_oracle_success": 0.6666666666666666,
2677
+ "hidden_chart_oracle_utility": 1.4079964955647786,
2678
+ "outcome_ptr": 0.6666666666666666,
2679
+ "predicted_margin": 0.32087827260133633,
2680
+ "proposal_oracle_success": 0.6666666666666666,
2681
+ "proposal_oracle_utility": 1.349846323331197,
2682
+ "selected_candidate_index": 3.0,
2683
+ "selected_success": 0.0,
2684
+ "selected_success_gain_over_base": 0.0,
2685
+ "selected_utility": 0.09965786089499791,
2686
+ "selected_utility_gain_over_base": -0.017935288449128468,
2687
+ "selector_regret": 1.2501884624361992,
2688
+ "success_selector_gap": 0.6666666666666666,
2689
+ "success_support_gap": 0.0,
2690
+ "support_gap": 0.05815017223358154,
2691
+ "tau": 0.27486681252026457
2692
+ },
2693
+ "767747424/train0": {
2694
+ "base_success": 0.25,
2695
+ "base_utility": 0.5731993420049548,
2696
+ "coverage": 0.5,
2697
+ "execute_generated": 0.5,
2698
+ "fallback_rate": 0.5,
2699
+ "hidden_chart_oracle_success": 0.25,
2700
+ "hidden_chart_oracle_utility": 0.6737305596470833,
2701
+ "outcome_ptr": 0.5,
2702
+ "predicted_margin": 0.26524823731305763,
2703
+ "proposal_oracle_success": 0.5,
2704
+ "proposal_oracle_utility": 1.0424291878007352,
2705
+ "selected_candidate_index": 4.0,
2706
+ "selected_success": 0.0,
2707
+ "selected_success_gain_over_base": -0.25,
2708
+ "selected_utility": 0.07187687387340702,
2709
+ "selected_utility_gain_over_base": -0.5013224681315478,
2710
+ "selector_regret": 0.9792825908807572,
2711
+ "success_selector_gap": 0.5,
2712
+ "success_support_gap": 0.0,
2713
+ "support_gap": 0.06562284985557199,
2714
+ "tau": 0.27486681252026457
2715
+ },
2716
+ "767747424/train1": {
2717
+ "base_success": 0.25,
2718
+ "base_utility": 0.5731993420049548,
2719
+ "coverage": 0.0,
2720
+ "execute_generated": 0.0,
2721
+ "fallback_rate": 1.0,
2722
+ "hidden_chart_oracle_success": 0.25,
2723
+ "hidden_chart_oracle_utility": 0.6737305596470833,
2724
+ "outcome_ptr": 0.25,
2725
+ "predicted_margin": 0.24280759033280117,
2726
+ "proposal_oracle_success": 0.25,
2727
+ "proposal_oracle_utility": 0.5951075805351138,
2728
+ "selected_candidate_index": 1.75,
2729
+ "selected_success": 0.25,
2730
+ "selected_success_gain_over_base": 0.0,
2731
+ "selected_utility": 0.5731993420049548,
2732
+ "selected_utility_gain_over_base": 0.0,
2733
+ "selector_regret": 0.03301048465073109,
2734
+ "success_selector_gap": 0.0,
2735
+ "success_support_gap": 0.25,
2736
+ "support_gap": 0.5129444571211934,
2737
+ "tau": 0.27486681252026457
2738
+ },
2739
+ "767747424/train2": {
2740
+ "base_success": 0.25,
2741
+ "base_utility": 0.5731993420049548,
2742
+ "coverage": 0.25,
2743
+ "execute_generated": 0.25,
2744
+ "fallback_rate": 0.75,
2745
+ "hidden_chart_oracle_success": 0.25,
2746
+ "hidden_chart_oracle_utility": 0.6737305596470833,
2747
+ "outcome_ptr": 0.5,
2748
+ "predicted_margin": 0.2516374530397714,
2749
+ "proposal_oracle_success": 0.25,
2750
+ "proposal_oracle_utility": 0.6134321661666036,
2751
+ "selected_candidate_index": 3.25,
2752
+ "selected_success": 0.25,
2753
+ "selected_success_gain_over_base": 0.0,
2754
+ "selected_utility": 0.5817436082288623,
2755
+ "selected_utility_gain_over_base": 0.00854426622390747,
2756
+ "selector_regret": 0.033712657168507576,
2757
+ "success_selector_gap": 0.0,
2758
+ "success_support_gap": 0.25,
2759
+ "support_gap": 0.49461987148970366,
2760
+ "tau": 0.27486681252026457
2761
+ }
2762
+ }
workspace/runs/ctt_base_context_obs_learned_dominance_context_success_envclip_train_to_test/metrics_by_task.json ADDED
@@ -0,0 +1,117 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "LiftPegUpright-v1": {
3
+ "base_success": 0.2,
4
+ "base_utility": 0.9498705387115478,
5
+ "coverage": 0.26666666666666666,
6
+ "execute_generated": 0.26666666666666666,
7
+ "fallback_rate": 0.7333333333333333,
8
+ "hidden_chart_oracle_success": 1.0,
9
+ "hidden_chart_oracle_utility": 2.0,
10
+ "outcome_ptr": 0.4666666666666667,
11
+ "predicted_margin": 0.2629060938740916,
12
+ "proposal_oracle_success": 0.4666666666666667,
13
+ "proposal_oracle_utility": 1.2244940986235937,
14
+ "selected_candidate_index": 3.2666666666666666,
15
+ "selected_success": 0.26666666666666666,
16
+ "selected_success_gain_over_base": 0.06666666666666667,
17
+ "selected_utility": 0.9406938781340917,
18
+ "selected_utility_gain_over_base": -0.009176660577456157,
19
+ "selector_regret": 0.39854816198348997,
20
+ "success_selector_gap": 0.26666666666666666,
21
+ "success_support_gap": 0.5333333333333333,
22
+ "support_gap": 0.7755059013764064,
23
+ "tau": 0.2748668125202646
24
+ },
25
+ "PickCube-v1": {
26
+ "base_success": 0.0,
27
+ "base_utility": 0.5839591026306152,
28
+ "coverage": 0.3333333333333333,
29
+ "execute_generated": 0.3333333333333333,
30
+ "fallback_rate": 0.6666666666666666,
31
+ "hidden_chart_oracle_success": 1.0,
32
+ "hidden_chart_oracle_utility": 2.0,
33
+ "outcome_ptr": 0.0,
34
+ "predicted_margin": 0.2631532926043247,
35
+ "proposal_oracle_success": 0.0,
36
+ "proposal_oracle_utility": 0.13425676027933756,
37
+ "selected_candidate_index": 2.0,
38
+ "selected_success": 0.0,
39
+ "selected_success_gain_over_base": 0.0,
40
+ "selected_utility": 0.41766610989967984,
41
+ "selected_utility_gain_over_base": -0.16629299273093542,
42
+ "selector_regret": 0.03272441774606705,
43
+ "success_selector_gap": 0.0,
44
+ "success_support_gap": 1.0,
45
+ "support_gap": 1.8657432397206624,
46
+ "tau": 0.27486681252026457
47
+ },
48
+ "PullCube-v1": {
49
+ "base_success": 0.36,
50
+ "base_utility": 0.7751704612234608,
51
+ "coverage": 0.4666666666666667,
52
+ "execute_generated": 0.4666666666666667,
53
+ "fallback_rate": 0.5333333333333333,
54
+ "hidden_chart_oracle_success": 0.52,
55
+ "hidden_chart_oracle_utility": 1.1467878860235214,
56
+ "outcome_ptr": 0.44,
57
+ "predicted_margin": 0.28018273372980096,
58
+ "proposal_oracle_success": 0.4666666666666667,
59
+ "proposal_oracle_utility": 0.9816135169876119,
60
+ "selected_candidate_index": 2.96,
61
+ "selected_success": 0.28,
62
+ "selected_success_gain_over_base": -0.08,
63
+ "selected_utility": 0.6107603244219596,
64
+ "selected_utility_gain_over_base": -0.16441013680150113,
65
+ "selector_regret": 0.5260294480016455,
66
+ "success_selector_gap": 0.26666666666666666,
67
+ "success_support_gap": 0.09333333333333334,
68
+ "support_gap": 0.23527305972451965,
69
+ "tau": 0.2748668125202645
70
+ },
71
+ "PushCube-v1": {
72
+ "base_success": 0.16666666666666666,
73
+ "base_utility": 0.3924041787783305,
74
+ "coverage": 1.0,
75
+ "execute_generated": 1.0,
76
+ "fallback_rate": 0.0,
77
+ "hidden_chart_oracle_success": 0.8333333333333334,
78
+ "hidden_chart_oracle_utility": 1.7015800600250561,
79
+ "outcome_ptr": 0.7777777777777778,
80
+ "predicted_margin": 0.5029348491622321,
81
+ "proposal_oracle_success": 0.6666666666666666,
82
+ "proposal_oracle_utility": 1.365773618221283,
83
+ "selected_candidate_index": 1.7222222222222223,
84
+ "selected_success": 0.4444444444444444,
85
+ "selected_success_gain_over_base": 0.2777777777777778,
86
+ "selected_utility": 0.9390517406993442,
87
+ "selected_utility_gain_over_base": 0.5466475619210137,
88
+ "selector_regret": 0.42672187752193874,
89
+ "success_selector_gap": 0.2222222222222222,
90
+ "success_support_gap": 0.3333333333333333,
91
+ "support_gap": 0.634226381778717,
92
+ "tau": 0.2748668125202646
93
+ },
94
+ "StackCube-v1": {
95
+ "base_success": 0.2727272727272727,
96
+ "base_utility": 0.7763947045261209,
97
+ "coverage": 0.7878787878787878,
98
+ "execute_generated": 0.7878787878787878,
99
+ "fallback_rate": 0.21212121212121213,
100
+ "hidden_chart_oracle_success": 1.0,
101
+ "hidden_chart_oracle_utility": 2.0,
102
+ "outcome_ptr": 0.5757575757575758,
103
+ "predicted_margin": 0.29881356602809567,
104
+ "proposal_oracle_success": 0.45454545454545453,
105
+ "proposal_oracle_utility": 1.0546941955884297,
106
+ "selected_candidate_index": 3.0,
107
+ "selected_success": 0.18181818181818182,
108
+ "selected_success_gain_over_base": -0.09090909090909091,
109
+ "selected_utility": 0.5421693889480649,
110
+ "selected_utility_gain_over_base": -0.2342253155780561,
111
+ "selector_regret": 0.5153396707592588,
112
+ "success_selector_gap": 0.2727272727272727,
113
+ "success_support_gap": 0.5454545454545454,
114
+ "support_gap": 0.9453058044115702,
115
+ "tau": 0.27486681252026457
116
+ }
117
+ }
workspace/runs/ctt_base_context_obs_learned_dominance_context_success_envclip_train_to_test/report.md ADDED
@@ -0,0 +1,17 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ # Learned Dominance-Calibrated CTT Selector
2
+
3
+ Calibration rows: `432`
4
+ Eval rows: `144`
5
+ Selected ridge lambda: `100.0`
6
+ Tau: `0.274867`
7
+ Feature set: `context`
8
+ Target: `success`
9
+
10
+ The ridge calibrator and threshold are fit on calibration measured rows only. Eval outcomes are used only for reporting.
11
+
12
+ | Split | Coverage | Fallback | Base success | Selected success | Proposal oracle | OutcomePTR | Success support gap | Success selector gap |
13
+ | --- | ---: | ---: | ---: | ---: | ---: | ---: | ---: | ---: |
14
+ | calibration | 0.5741 | 0.4259 | 0.2778 | 0.2940 | 0.4722 | 0.4907 | 0.2315 | 0.2199 |
15
+ | eval | 0.5833 | 0.4167 | 0.2917 | 0.2708 | 0.4792 | 0.5069 | 0.2917 | 0.2569 |
16
+
17
+ This is a selector diagnostic over already measured candidates, not a new rollout.
workspace/runs/ctt_base_context_obs_learned_dominance_context_success_envclip_train_to_test/split_hash.txt ADDED
@@ -0,0 +1,14 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "calibration_target_index": {
3
+ "content_hash": "67f7b4a692f7c7e71da24378dc71e3399c5d35afc14c494869eba6087b765c42",
4
+ "retrieval_index_allowed": true,
5
+ "split": "train",
6
+ "split_hash": "867a94a1c67686c082765671a7bf8d74a9d633085e1b2619c83197c60bf58372"
7
+ },
8
+ "eval_target_index": {
9
+ "content_hash": "0c38d73dc9bc4ab756e12006671b5d12f1a23358b281e54b3f6a9ae3cbf6e9af",
10
+ "retrieval_index_allowed": false,
11
+ "split": "test",
12
+ "split_hash": "84151df9e88c219b76751677558cc5bbbc19082bf358cea7330158e9e403e25f"
13
+ }
14
+ }
workspace/runs/ctt_base_context_obs_learned_dominance_context_success_envclip_train_to_test/table.tex ADDED
@@ -0,0 +1,8 @@
 
 
 
 
 
 
 
 
 
1
+ % Auto-generated by scripts/eval_learned_dominance_selector.py
2
+ \begin{tabular}{lrrrrrrrr}
3
+ \toprule
4
+ Rows & Coverage & Fallback & Base succ. & Selected succ. & Oracle succ. & OutcomePTR & Succ. support gap & Succ. selector gap \\
5
+ \midrule
6
+ 144 & 0.5833 & 0.4167 & 0.2917 & 0.2708 & 0.4792 & 0.5069 & 0.2917 & 0.2569 \\
7
+ \bottomrule
8
+ \end{tabular}
workspace/runs/ctt_base_context_obs_learned_dominance_context_success_envclip_train_to_test/train.log ADDED
@@ -0,0 +1,4 @@
 
 
 
 
 
1
+ trained ridge dominance calibrator on calibration measured rows only
2
+ selected_lambda=100.0
3
+ tau=0.274867
4
+ calibration_input=runs/ctt_base_context_obs_train_cal_envclip_rollout_comparison/combined_measured_candidates.json
workspace/runs/ctt_base_context_obs_learned_dominance_context_tangent_envclip_train_to_test/command.txt ADDED
@@ -0,0 +1 @@
 
 
1
+ python scripts/eval_learned_dominance_selector.py --calibration-input runs/ctt_base_context_obs_train_cal_envclip_rollout_comparison/combined_measured_candidates.json --calibration-target-index data/cil_charts_rgb_refs/train/index.json --eval-input runs/ctt_base_context_obs_test_envclip_rollout_comparison/combined_measured_candidates.json --eval-target-index data/cil_charts_rgb_refs/test/index.json --checkpoint-template runs/ctt_residual_base_context_obs_seed{seed}/model.pt --out-dir runs/ctt_base_context_obs_learned_dominance_context_tangent_envclip_train_to_test --feature-set context_tangent --target utility_margin --k 8 --bootstrap-samples 1000
workspace/runs/ctt_base_context_obs_learned_dominance_context_tangent_envclip_train_to_test/config.yaml ADDED
@@ -0,0 +1,11 @@
 
 
 
 
 
 
 
 
 
 
 
 
1
+ bootstrap_samples: 1000
2
+ calibration_input: runs/ctt_base_context_obs_train_cal_envclip_rollout_comparison/combined_measured_candidates.json
3
+ calibration_target_index: data/cil_charts_rgb_refs/train/index.json
4
+ checkpoint_template: runs/ctt_residual_base_context_obs_seed{seed}/model.pt
5
+ eval_input: runs/ctt_base_context_obs_test_envclip_rollout_comparison/combined_measured_candidates.json
6
+ eval_target_index: data/cil_charts_rgb_refs/test/index.json
7
+ feature_set: context_tangent
8
+ k: 8
9
+ out_dir: runs/ctt_base_context_obs_learned_dominance_context_tangent_envclip_train_to_test
10
+ ridge_lambdas: 0,0.01,0.1,1,10,100
11
+ target: utility_margin