List your company

Science Question Answering on SciQ (Score, Delta (Δ), Avg. Rank)

98.53ScoreQwen2.5(14B)T
78.218883.491988.76594.0381Jul 4, 2026

Evaluation results

MethodLinks
Qwen2.5(14B)TEvaluation Protocol=Generation-based2026.0798.53——
RG-OPDEvaluation Protocol=Generation-based2026.0797.81.072
Reverse KLEvaluation Protocol=Generation-based2026.0797.370.643
TSD-KDEvaluation Protocol=Generation-based2026.0797.20.472.6
Qwen2.5(1.5B)SEvaluation Protocol=Generation-based2026.0796.7302.4
TSD-KDEvaluation Protocol=Likelihood-based2026.0793.64.12.6
Qwen2.5(1.5B)SEvaluation Protocol=Likelihood-based2026.0789.502.4
Qwen2.5(14B)TEvaluation Protocol=Likelihood-based2026.0786.8——
Reverse KLEvaluation Protocol=Likelihood-based2026.0780-9.53
RG-OPDEvaluation Protocol=Likelihood-based2026.0779-10.52