Correlation Evaluation on 37k Questions Dataset
0.2379Pearson's rhoLLM-as-a-Judge
Evaluation results
| Method | Links | ||
|---|---|---|---|
| LLM-as-a-JudgeJudge/Expert Model Size=7B2026.01 | 0.2379 | — | |
| Peer PredictionJudge/Expert Model Size=0.5B2026.01 | 0.1947 | — | |
| LLM-as-a-JudgeJudge/Expert Model Size=3B2026.01 | 0.1852 | — | |
| ConsistencyJudge/Expert Model Size=7B2026.01 | 0.151 | — | |
| LLM-as-a-JudgeJudge/Expert Model Size=1B2026.01 | 0.1197 | — | |
| Peer PredictionJudge/Expert Model Size=135M2026.01 | 0.0986 | — | |
| Peer PredictionJudge/Expert Model Size=360M2026.01 | 0.0981 | — | |
| ConsistencyJudge/Expert Model Size=0.5B2026.01 | 0.0801 | — | |
| ConsistencyJudge/Expert Model Size=3B2026.01 | 0.0785 | — | |
| ConsistencyJudge/Expert Model Size=1B2026.01 | 0.0536 | — | |
| Peer PredictionJudge/Expert Model Size=1B2026.01 | 0.0462 | — | |
| ConsistencyJudge/Expert Model Size=135M2026.01 | 0.0408 | — | |
| LLM-as-a-JudgeJudge/Expert Model Size=135M2026.01 | 0.0352 | — | |
| LLM-as-a-JudgeJudge/Expert Model Size=360M2026.01 | -0.0142 | — | |
| ConsistencyJudge/Expert Model Size=360M2026.01 | -0.0229 | — | |
| LLM-as-a-JudgeJudge/Expert Model Size=0.5B2026.01 | -0.0236 | — | |
| Peer PredictionJudge/Expert Model Size=3B2026.01 | -0.0421 | — | |
| Peer PredictionJudge/Expert Model Size=7B2026.01 | -0.1595 | — |