model large_stringclasses 1
value | id large_stringlengths 10 30 | s int64 0 99 | split large_stringclasses 4
values | valid bool 2
classes |
|---|---|---|---|---|
qwen3-8b | irrelevance_100 | 0 | train | true |
qwen3-8b | irrelevance_100 | 8 | train | true |
qwen3-8b | irrelevance_100 | 16 | train | true |
qwen3-8b | irrelevance_100 | 24 | train | true |
qwen3-8b | irrelevance_100 | 32 | train | true |
qwen3-8b | irrelevance_100 | 40 | train | true |
qwen3-8b | irrelevance_100 | 48 | train | true |
qwen3-8b | irrelevance_100 | 56 | train | true |
qwen3-8b | irrelevance_100 | 64 | train | true |
qwen3-8b | irrelevance_100 | 72 | train | true |
qwen3-8b | irrelevance_100 | 80 | train | true |
qwen3-8b | irrelevance_100 | 88 | train | true |
qwen3-8b | irrelevance_100 | 96 | train | true |
qwen3-8b | irrelevance_101 | 4 | test | false |
qwen3-8b | irrelevance_101 | 12 | test | false |
qwen3-8b | irrelevance_101 | 20 | test | true |
qwen3-8b | irrelevance_101 | 28 | test | true |
qwen3-8b | irrelevance_101 | 36 | test | true |
qwen3-8b | irrelevance_101 | 44 | test | true |
qwen3-8b | irrelevance_101 | 52 | test | false |
qwen3-8b | irrelevance_101 | 60 | test | false |
qwen3-8b | irrelevance_101 | 68 | test | true |
qwen3-8b | irrelevance_101 | 76 | test | false |
qwen3-8b | irrelevance_101 | 84 | test | false |
qwen3-8b | irrelevance_101 | 92 | test | true |
qwen3-8b | irrelevance_102 | 0 | test | true |
qwen3-8b | irrelevance_102 | 8 | test | true |
qwen3-8b | irrelevance_102 | 16 | test | true |
qwen3-8b | irrelevance_102 | 24 | test | true |
qwen3-8b | irrelevance_102 | 32 | test | true |
qwen3-8b | irrelevance_102 | 40 | test | true |
qwen3-8b | irrelevance_102 | 48 | test | true |
qwen3-8b | irrelevance_102 | 56 | test | true |
qwen3-8b | irrelevance_102 | 64 | test | true |
qwen3-8b | irrelevance_102 | 72 | test | true |
qwen3-8b | irrelevance_102 | 80 | test | true |
qwen3-8b | irrelevance_102 | 88 | test | true |
qwen3-8b | irrelevance_102 | 96 | test | true |
qwen3-8b | irrelevance_104 | 4 | train | true |
qwen3-8b | irrelevance_104 | 12 | train | true |
qwen3-8b | irrelevance_104 | 20 | train | true |
qwen3-8b | irrelevance_104 | 28 | train | true |
qwen3-8b | irrelevance_104 | 36 | train | true |
qwen3-8b | irrelevance_104 | 44 | train | true |
qwen3-8b | irrelevance_104 | 52 | train | true |
qwen3-8b | irrelevance_104 | 60 | train | true |
qwen3-8b | irrelevance_104 | 68 | train | true |
qwen3-8b | irrelevance_104 | 76 | train | true |
qwen3-8b | irrelevance_104 | 84 | train | true |
qwen3-8b | irrelevance_104 | 92 | train | true |
qwen3-8b | irrelevance_105 | 0 | train | true |
qwen3-8b | irrelevance_105 | 8 | train | true |
qwen3-8b | irrelevance_105 | 16 | train | true |
qwen3-8b | irrelevance_105 | 24 | train | true |
qwen3-8b | irrelevance_105 | 32 | train | true |
qwen3-8b | irrelevance_105 | 40 | train | true |
qwen3-8b | irrelevance_105 | 48 | train | true |
qwen3-8b | irrelevance_105 | 56 | train | true |
qwen3-8b | irrelevance_105 | 64 | train | true |
qwen3-8b | irrelevance_105 | 72 | train | true |
qwen3-8b | irrelevance_105 | 80 | train | true |
qwen3-8b | irrelevance_105 | 88 | train | true |
qwen3-8b | irrelevance_105 | 96 | train | true |
qwen3-8b | irrelevance_107 | 4 | train | true |
qwen3-8b | irrelevance_107 | 12 | train | true |
qwen3-8b | irrelevance_107 | 20 | train | true |
qwen3-8b | irrelevance_107 | 28 | train | true |
qwen3-8b | irrelevance_107 | 36 | train | true |
qwen3-8b | irrelevance_107 | 44 | train | true |
qwen3-8b | irrelevance_107 | 52 | train | true |
qwen3-8b | irrelevance_107 | 60 | train | true |
qwen3-8b | irrelevance_107 | 68 | train | true |
qwen3-8b | irrelevance_107 | 76 | train | true |
qwen3-8b | irrelevance_107 | 84 | train | true |
qwen3-8b | irrelevance_107 | 92 | train | true |
qwen3-8b | irrelevance_109 | 0 | val | true |
qwen3-8b | irrelevance_109 | 8 | val | true |
qwen3-8b | irrelevance_109 | 16 | val | true |
qwen3-8b | irrelevance_109 | 24 | val | true |
qwen3-8b | irrelevance_109 | 32 | val | true |
qwen3-8b | irrelevance_109 | 40 | val | true |
qwen3-8b | irrelevance_109 | 48 | val | true |
qwen3-8b | irrelevance_109 | 56 | val | true |
qwen3-8b | irrelevance_109 | 64 | val | true |
qwen3-8b | irrelevance_109 | 72 | val | true |
qwen3-8b | irrelevance_109 | 80 | val | true |
qwen3-8b | irrelevance_109 | 88 | val | true |
qwen3-8b | irrelevance_109 | 96 | val | true |
qwen3-8b | irrelevance_11 | 4 | train | false |
qwen3-8b | irrelevance_11 | 12 | train | true |
qwen3-8b | irrelevance_11 | 20 | train | true |
qwen3-8b | irrelevance_11 | 28 | train | false |
qwen3-8b | irrelevance_11 | 36 | train | false |
qwen3-8b | irrelevance_11 | 44 | train | false |
qwen3-8b | irrelevance_11 | 52 | train | false |
qwen3-8b | irrelevance_11 | 60 | train | false |
qwen3-8b | irrelevance_11 | 68 | train | false |
qwen3-8b | irrelevance_11 | 76 | train | false |
qwen3-8b | irrelevance_11 | 84 | train | true |
qwen3-8b | irrelevance_11 | 92 | train | false |
End of preview. Expand in Data Studio
YAML Metadata Warning:empty or missing yaml metadata in repo card
Check out the documentation for more information.
BFCL probe activations, answer-INCLUSIVE (prompt+think+answer prefill), qwen3-8b, dense layers 10-34, last-token + mean pooling. acts_full_output/ = 1,800-problem split, acts_full_output_extra/ = 1,841 complement. kfold_ansinc/ = K=3 answer-inclusive linear classifiers (layer 28) deployed in live best-of-100 (condition C). Rebuild: experiments/bfcl_cot_clf/ in github.com/genlm/rollouts (clement/wip).
- Downloads last month
- 80