Best-of-N evaluation on RMB
59.69AccuracyPC2-based LLM-as-a-Judge
Evaluation Results
| Method | Links | |
|---|---|---|
| PC2-based LLM-as-a-JudgeEvaluation Method=ours2025.05 | 59.69 | |
| Naive Pointwise EvaluationEvaluation Method=naive2025.05 | 40.68 |
Loading the SOTA2 catalog…
| Method | Links | |
|---|---|---|
| PC2-based LLM-as-a-JudgeEvaluation Method=ours2025.05 | 59.69 | |
| Naive Pointwise EvaluationEvaluation Method=naive2025.05 | 40.68 |