List your company

Uncertainty Estimation on HumanEval-X C++ (test)

0.649AUROCMax Entropy
0.517960.551980.5860.62002Jun 8, 2026

Evaluation results

MethodLinks
Max EntropyModel=Qwen3-Coder-30B2026.060.649
Top-5 EntropyModel=Qwen3-Coder-30B2026.060.648
Consistency (BLEU)Model=Qwen3-Coder-30B2026.060.581
Consistency (VR)Model=Qwen3-Coder-30B2026.060.572
Mean EntropyModel=Qwen3-Coder-30B2026.060.543
Mean ProbabilityModel=Qwen3-Coder-30B2026.060.523