Instruction Following on AlpacaEval LC 2.0
78.61AlpacaEval 2.0 LC ScoreGaussian time sampling
Evaluation Results
| Method | Links | ||
|---|---|---|---|
| Gaussian time samplingJudge=GPT-4o, Evaluation Protocol=Averaged over 5 random seeds2026.05 | 78.61 | — | |
| Llama-3.1-Nemotron-70B-InstructFeedback + Edit Protocol=true2025.03 | 62.8 | 1.3 | |
| Llama-3.1-Nemotron-70B-InstructFeedback + Edit Protocol=false2025.03 | 57.6 | 1.65 | |
| GPT-4o-2024-05-132025.03 | 57.5 | 1.47 | |
| Claude-3-5-Sonnet-202406202025.03 | 52.4 | 1.47 | |
| Llama-3.1-405B-Instruct2025.03 | 39.3 | 1.43 | |
| Llama-3.1-70B-Instruct2025.03 | 38.1 | 0.9 | |
| Llama-3.3-70B-InstructFeedback + Edit Protocol=true2025.03 | 36.9 | 1.5 | |
| Llama-3.3-70B-InstructFeedback + Edit Protocol=false2025.03 | 35 | 1.45 |