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

Reasoning on Downstream Tasks (PiQA, Arc E, Arc C, HS, WG, BoolQ)

80.5PiQA Accuracy

FP16

50.3458.176673.83Apr 9, 2026
Updated 9d ago

Evaluation Results

MethodLinks
2026.04
80.577.549.279.472.380.673.3
2026.04
7974.546.3766977.770.4
2026.04
7974.546.3766977.770.4
2026.04
62.845.526.444.155.862.349.5
2026.04
5939.524.940.352.661.446.3
2026.04
57.735.523.6355356.643.8
2026.04
57.339.524.435.752.257.844.4
2026.04
57.134.923.633.253.761.444
2026.04
51.525.525.633.151.757.640.8