Counterfactual Subset-Utility Evaluation on OpenBookQA
0.35Spearman CorrelationGRASP
Evaluation Results
| Method | Links | ||
|---|---|---|---|
| GRASPEvaluation Protocol=Linear Datamodeling Score (LDS), Model Architecture=50M-parameter causal decoder-only Transformer2026.06 | 0.35 | 0.855 | |
| TRAKEvaluation Protocol=Linear Datamodeling Score (LDS), Model Architecture=50M-parameter causal decoder-only Transformer2026.06 | 0.182 | 0.582 | |
| TracInEvaluation Protocol=Linear Datamodeling Score (LDS), Model Architecture=50M-parameter causal decoder-only Transformer2026.06 | 0.151 | 0.578 | |
| BM25Evaluation Protocol=Linear Datamodeling Score (LDS), Model Architecture=50M-parameter causal decoder-only Transformer2026.06 | 0.064 | 0.508 |