List your company

Counterfactual Subset-Utility Evaluation on SciQ

0.424Spearman Correlation (ρ)GRASP
0.177520.241510.30550.36949Jun 5, 2026

Evaluation results

MethodLinks
GRASPEvaluation Protocol=Linear Datamodeling Score (LDS), Model Architecture=50M-parameter causal decoder-only Transformer2026.060.42490.6
TRAKEvaluation Protocol=Linear Datamodeling Score (LDS), Model Architecture=50M-parameter causal decoder-only Transformer2026.060.24264.1
TracInEvaluation Protocol=Linear Datamodeling Score (LDS), Model Architecture=50M-parameter causal decoder-only Transformer2026.060.22360.9
BM25Evaluation Protocol=Linear Datamodeling Score (LDS), Model Architecture=50M-parameter causal decoder-only Transformer2026.060.18760.9