LLM Re-ranking on AlpacaEval (GPT4-1106-Preview Baseline)
32.44Win Rate (LC)TRLM-Ba
Evaluation results
| Method | Links | |||||||
|---|---|---|---|---|---|---|---|---|
| TRLM-BaInference Style=Response -> Query2024.12 | 32.44 | 24.35 | 24.04 | 1.27 | 192 | 610 | 3 | |
| TRLM-FoBa (backward)Inference Style=Response -> Query2024.12 | 31.18 | 22.72 | 21.99 | 1.24 | 176 | 627 | 2 | |
| TRLM-FoBa (forward)Inference Style=Response -> Query2024.12 | 30.55 | 22.85 | 22.48 | 1.25 | 180 | 623 | 2 | |
| TRLM-FoInference Style=Response -> Query2024.12 | 29.19 | 22.68 | 21.3 | 1.24 | 170 | 632 | 3 | |
| SelfInference Style=Query -> Response2024.12 | 27.05 | 17.66 | 17.14 | 1.15 | 136 | 665 | 4 | |
| One Generation2024.12 | 24.38 | 18.18 | 17.08 | 1.16 | 135 | 665 | 5 | |
| Forward BaselineInference Style=Query -> Response2024.12 | 24.27 | 17.13 | 15.78 | 1.12 | 126 | 677 | 2 |