Question Answering on OpenBookQA published (test)
92.6AccuracyJF-HPO
Evaluation Results
| Method | Links | |||||
|---|---|---|---|---|---|---|
| JF-HPOModel=Qwen-3 14B2026.06 | 92.6 | — | — | — | — | |
| BOHBModel=Qwen-2.5 7B2026.06 | 91 | — | — | — | — | |
| JF-HPOModel=Qwen-2.5 7B2026.06 | 91 | — | — | — | — | |
| VeRL RecipeModel=Qwen-3 14B2026.06 | 90.6 | — | — | — | — | |
| Random SearchModel=Qwen-2.5 7B2026.06 | 89.4 | — | — | — | — | |
| VeRL RecipeModel=Qwen-2.5 7B2026.06 | 88.2 | — | — | — | — | |
| JF-HPOModel=LLaMA-3.1 8B2026.06 | 87.8 | — | — | — | — | |
| BOHBModel=LLaMA-3.1 8B2026.06 | 85.8 | — | — | — | — | |
| Random SearchModel=LLaMA-3.1 8B2026.06 | 85.6 | — | — | — | — | |
| VeRL RecipeModel=LLaMA-3.1 8B2026.06 | 83 | — | — | — | — | |
| CORALBackbone=Qwen2.5-7B-Instruct, Transfer Experiments=Probe 12026.02 | 65.4 | 7.21 | 7.44 | 66.11 | 20.92 | |
| CORALBackbone=Mistral-7B-Instruct-v0.3, Transfer Experiments=Probe 12026.02 | 63.2 | 10.66 | 6.65 | 69.57 | 20.7 | |
| CORALBackbone=Deepseek-7B-Chat, Transfer Experiments=Probe 12026.02 | 62.8 | 11.47 | 8.19 | 81.25 | 23.19 | |
| CCPSBackbone=Mistral-7B-Instruct-v0.3, Transfer Experiments=Probe 12026.02 | 52 | 28.92 | 8.1 | 92.09 | 32.97 | |
| FewshotBackbone=Mistral-7B-Instruct-v0.3, Transfer Experiments=Probe 12026.02 | 50.4 | 11.4 | 8.8 | 70.18 | 24.5 | |
| ITIBackbone=Mistral-7B-Instruct-v0.3, Transfer Experiments=Probe 12026.02 | 50.23 | 11.4 | 7.91 | 60.5 | 22.48 | |
| FewshotBackbone=Qwen2.5-7B-Instruct, Transfer Experiments=Probe 12026.02 | 49.8 | 17.64 | 10.37 | 74.54 | 25.4 | |
| ITIBackbone=Qwen2.5-7B-Instruct, Transfer Experiments=Probe 12026.02 | 49.8 | 17.64 | 9.32 | 64.26 | 23.31 | |
| FewshotBackbone=Deepseek-7B-Chat, Transfer Experiments=Probe 12026.02 | 48 | 17.12 | 11.55 | 85.11 | 24.87 | |
| CCPSBackbone=Qwen2.5-7B-Instruct, Transfer Experiments=Probe 12026.02 | 47.6 | 15.13 | 10.46 | 72.6 | 26.53 | |
| CCPSBackbone=Deepseek-7B-Chat, Transfer Experiments=Probe 12026.02 | 47.6 | 19.57 | 11.43 | 80.47 | 28.73 | |
| ITIBackbone=Deepseek-7B-Chat, Transfer Experiments=Probe 12026.02 | 46.48 | 17.12 | 10.38 | 73.37 | 22.82 | |
| SteerConfBackbone=Mistral-7B-Instruct-v0.3, Transfer Experiments=Probe 12026.02 | 34.44 | 17.41 | 11.25 | 69.03 | 30.03 | |
| SteerConfBackbone=Qwen2.5-7B-Instruct, Transfer Experiments=Probe 12026.02 | 34.03 | 26.95 | 13.26 | 73.32 | 31.13 | |
| SteerConfBackbone=Deepseek-7B-Chat, Transfer Experiments=Probe 12026.02 | 32.8 | 26.15 | 14.76 | 83.71 | 30.48 |