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

Question Answering on GPQA (Accuracy, Consistency, # traces)

2,513# Traces

PETS-Offline

2,149.244,604.627,0609,515.38Feb 18, 2026
Updated 1mo ago

Evaluation Results

MethodLinks
2026.02
2,51310076.3
2026.02
2,58010082.5
2026.02
2,60710071.8
2026.02
2,78010069.7
2026.02
2,8879977
2026.02
3,18010075.7
2026.02
3,5409982.1
2026.02
3,6679970.3
2026.02
3,6879972.1
2026.02
3,6939976.2
2026.02
9,9739676.5
2026.02
10,3679671.5
2026.02
10,3939781.5
2026.02
10,5009680.8
2026.02
10,8539675.1
2026.02
11,0139668.9
2026.02
11,1939675.6
2026.02
11,4009576.4
2026.02
11,4539569.9
2026.02
11,6079572.2