StreamSense-Serve-v4-Router / e2e_diagnostics.json
jatshi's picture
Publish verified StreamSense-Serve 4.0 router
4ff5147 verified
Raw History Blame Contribute Delete
2.74 kB
{
"schema_version": 1,
"bootstrap_unit": "group_id",
"bootstrap_iterations": 10000,
"independent_groups": 8,
"systems": {
"always_vlm": {
"cases": 32,
"quality_pass_rate": 0.6875,
"group_bootstrap_95_ci": [
0.46875,
0.875
],
"failure_counts": {
"state": 9,
"answer_fact": 1,
"citation": 4
},
"per_group_quality_pass_rate": {
"meeting-conflict-003": 0.75,
"meeting-conflict-005": 0.25,
"meeting-normal-001": 1.0,
"meeting-visual-001": 1.0,
"meeting-visual-008": 1.0,
"safety-risk-003": 0.25,
"safety-risk-004": 0.5,
"support-normal-001": 0.75
}
},
"learned": {
"cases": 32,
"quality_pass_rate": 0.65625,
"group_bootstrap_95_ci": [
0.40625,
0.875
],
"failure_counts": {
"state": 10,
"answer_fact": 1,
"citation": 5
},
"per_group_quality_pass_rate": {
"meeting-conflict-003": 0.75,
"meeting-conflict-005": 0.0,
"meeting-normal-001": 1.0,
"meeting-visual-001": 1.0,
"meeting-visual-008": 1.0,
"safety-risk-003": 0.25,
"safety-risk-004": 0.5,
"support-normal-001": 0.75
}
},
"never_vlm": {
"cases": 32,
"quality_pass_rate": 0.78125,
"group_bootstrap_95_ci": [
0.53125,
0.96875
],
"failure_counts": {
"state": 6,
"answer_fact": 1,
"citation": 5
},
"per_group_quality_pass_rate": {
"meeting-conflict-003": 0.75,
"meeting-conflict-005": 0.0,
"meeting-normal-001": 1.0,
"meeting-visual-001": 1.0,
"meeting-visual-008": 1.0,
"safety-risk-003": 0.75,
"safety-risk-004": 1.0,
"support-normal-001": 0.75
}
}
},
"paired_quality": {
"always_vlm_minus_learned": {
"paired_cases": 32,
"quality_rate_delta": 0.03125,
"group_bootstrap_95_ci": [
0.0,
0.09375
],
"both_pass": 21,
"left_only_pass": 1,
"right_only_pass": 0,
"both_fail": 10
},
"always_vlm_minus_never_vlm": {
"paired_cases": 32,
"quality_rate_delta": -0.09375,
"group_bootstrap_95_ci": [
-0.25,
0.0625
],
"both_pass": 21,
"left_only_pass": 1,
"right_only_pass": 4,
"both_fail": 6
},
"learned_minus_never_vlm": {
"paired_cases": 32,
"quality_rate_delta": -0.125,
"group_bootstrap_95_ci": [
-0.3125,
0.0
],
"both_pass": 21,
"left_only_pass": 0,
"right_only_pass": 4,
"both_fail": 7
}
}
}