Mathematical Reasoning on MATH (AUROC, AUARC)
0.7121AUROCMATU
Evaluation Results
| Method | Links | ||
|---|---|---|---|
| MATULLM=Llama3.1-8B2026.04 | 0.7121 | 0.6518 | |
| MATULLM=Qwen2.5-7B2026.04 | 0.6966 | 0.8585 | |
| MATULLM=GPT-4o2026.04 | 0.6675 | 0.6439 | |
| P(true)LLM=Qwen2.5-7B2026.04 | 0.6602 | 0.8291 | |
| P(true)LLM=Llama3.1-8B2026.04 | 0.6581 | 0.6225 | |
| Eigv(Agre)-finalLLM=Qwen2.5-7B2026.04 | 0.6506 | 0.8212 | |
| SAUP-MultipleLLM=Llama3.1-8B2026.04 | 0.6477 | 0.6065 | |
| SAUP-MultipleLLM=Qwen2.5-7B2026.04 | 0.6396 | 0.8119 | |
| Eigv(Agre)-finalLLM=GPT-4o2026.04 | 0.6359 | 0.6133 | |
| Eigv(Agre)-finalLLM=Llama3.1-8B2026.04 | 0.634 | 0.6059 | |
| SAUP-SingleLLM=Llama3.1-8B2026.04 | 0.6339 | 0.6008 | |
| Eigv(Agre)-WholeLLM=Qwen2.5-7B2026.04 | 0.6314 | 0.8081 | |
| Eigv(Agre)-WholeLLM=GPT-4o2026.04 | 0.6308 | 0.6115 | |
| SAUP-SingleLLM=Qwen2.5-7B2026.04 | 0.6261 | 0.7982 | |
| P(true)LLM=GPT-4o2026.04 | 0.6226 | 0.607 | |
| Eigv(Agre)-WholeLLM=Llama3.1-8B2026.04 | 0.6215 | 0.5953 |