List your company

Uncertainty Calibration on GSM8k OOD (test)

3.15ECEEntropy-based (ours)
1.944810.079918.21526.3501Mar 6, 2026

Evaluation results

MethodLinks
Entropy-based (ours)Backbone=Qwen2.5-7B-Instruct, Reinforcement learning=GRPO, Reward design=Calibrated entropy-based reward signal2026.033.1566.73
Base+CoTBackbone=Qwen2.5-7B-Instruct, Reinforcement learning=None, Chain-of-thought reasoning=true2026.0322.2562.17
BaseBackbone=Qwen2.5-7B-Instruct, Reinforcement learning=None, Chain-of-thought reasoning=false2026.0332.2253.79
BrierBackbone=Qwen2.5-7B-Instruct, Reinforcement learning=GRPO, Reward design=Brier score2026.0333.2866.89