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

Scientific Question Answering on SciEval Static

81.52Accuracy

Reason, Reward, Refine

43.029653.022363.01573.0077Jul 6, 2026
Updated 18d ago

Evaluation Results

MethodLinks
2026.07
81.52
2026.07
79.29
2026.07
68.12
2026.07
68.09
2026.07
66.87
2026.07
65.19
2026.07
62.73
2026.07
62.43
2026.07
62.26
2026.07
61.59
2026.07
61.01
2026.07
59.32
2026.07
59.17
2026.07
57.93
2026.07
57.93
2026.07
57.37
2026.07
56.93
2026.07
56.87
2026.07
55.49
2026.07
44.51