LLM-as-a-judge on AlpacaEval
78.3CoverageHeuristic Selection
Evaluation results
| Method | Links | ||
|---|---|---|---|
| Heuristic SelectionCascade=M → L → O, Target agreement level=0.852026.05 | 78.3 | 0 | |
| Heuristic SelectionCascade=L → Q → O, Target agreement level=0.852026.05 | 74.5 | 0 | |
| CSE + PredictiveCascade=M → L → O, Target agreement level=0.852026.05 | 54.5 | 64.7 | |
| CSE + OursCascade=M → L → O, Target agreement level=0.852026.05 | 54.5 | 82.1 | |
| CSE + SimulatedCascade=M → L → O, Target agreement level=0.852026.05 | 52.6 | 72.8 | |
| CSE + VanillaCascade=M → L → O, Target agreement level=0.852026.05 | 52 | 75.7 | |
| CSE + RandomCascade=M → L → O, Target agreement level=0.852026.05 | 48.7 | 74.6 | |
| CSE + OursCascade=L → Q → O, Target agreement level=0.852026.05 | 38.4 | 94.8 | |
| CSE + PredictiveCascade=L → Q → O, Target agreement level=0.852026.05 | 37.6 | 86.9 | |
| CSE + VanillaCascade=L → Q → O, Target agreement level=0.852026.05 | 34.6 | 93.4 | |
| CSE + SimulatedCascade=L → Q → O, Target agreement level=0.852026.05 | 34.1 | 90.8 | |
| CSE + RandomCascade=L → Q → O, Target agreement level=0.852026.05 | 33.2 | 92 |