Share your thoughts, 1 month free Claude Pro on usSee more
WorkDL logo mark

Circuit Recovery Evaluation Suite (IOI, MCQA, Arithmetic, ARC) Llama-3 (3B to 8B) Transfer (test)

46Arithmetic (+) Score

NAP-IG-Inputs

24.1629.8335.541.17Apr 27, 2026
Updated 1mo ago

Evaluation Results

MethodLinks
2026.04
46631.0237979473
2026.04
46500.5453535652
2026.04
45490.2645303438
2026.04
40230.2740292831
2026.04
36--37282932
2026.04
33350.631596347
2026.04
29--67343441
2026.04
28440.559364343
2026.04
27570.4646272538
2026.04
27360.3226292730
2026.04
27--59565649
2026.04
25250.2525252525
2026.04
25310.3829575940