Preference Reconstruction on Research Questions
76.2Preference AccuracyDICAI
Evaluation results
| Method | Links | |
|---|---|---|
| DICAIJudge LLM=GPT-52026.06 | 76.2 | |
| ICAIJudge LLM=GPT-52026.06 | 71.8 | |
| AutoRubricJudge LLM=GPT-52026.06 | 58.76 | |
| CoTJudge LLM=GPT-52026.06 | 58.12 | |
| CoT-SCJudge LLM=GPT-52026.06 | 57.28 | |
| ToTJudge LLM=GPT-52026.06 | 56.96 | |
| Self-RefineJudge LLM=GPT-52026.06 | 54.87 |