List your company

Uncertainty Calibration on SciBench

78.7AUROCPE
54.98861.14467.373.456Jan 2, 2026

Evaluation results

MethodLinks
PEBackbone=Qwen2.5-7B-Instruct2026.0178.7
ET-PEBackbone=Qwen2.5-7B-Instruct2026.0178.5
p(True)Backbone=Qwen2.5-7B-Instruct2026.0177.1
LSBackbone=Qwen2.5-32B-Instruct2026.0176.8
ET-PEBackbone=Qwen2.5-32B-Instruct2026.0176.7
LSBackbone=Qwen2.5-7B-Instruct2026.0176.5
PEBackbone=Qwen2.5-32B-Instruct2026.0176.5
LSBackbone=Qwen2.5-14B-Instruct2026.0175.3
ET-PEBackbone=Qwen2.5-14B-Instruct2026.0175.1
PEBackbone=Qwen2.5-14B-Instruct2026.0173.7
p(True)Backbone=Qwen2.5-32B-Instruct2026.0170.8
SEBackbone=Qwen2.5-14B-Instruct2026.0163
SEBackbone=Qwen2.5-32B-Instruct2026.0163
LN-PEBackbone=Qwen2.5-14B-Instruct2026.0162.2
p(True)Backbone=Qwen2.5-14B-Instruct2026.0162
LN-PEBackbone=Qwen2.5-7B-Instruct2026.0158.9
SEBackbone=Qwen2.5-7B-Instruct2026.0157.7
LN-PEBackbone=Qwen2.5-32B-Instruct2026.0155.9