General Knowledge on GPQA
33.3AccuracyI-DPO + MaPPO
Evaluation Results
| Method | Links | |
|---|---|---|
| I-DPO + MaPPOModel=Qwen2.5-7B-Instruct2025.07 | 33.3 | |
| DPO + MaPPOModel=Qwen2.5-7B-Instruct2025.07 | 33.1 | |
| I-DPOModel=Qwen2.5-7B-Instruct2025.07 | 33 | |
| DPOModel=Qwen2.5-7B-Instruct2025.07 | 32 | |
| ITModel=Qwen2.5-7B-Instruct2025.07 | 31.5 | |
| ITModel=Llama-3-8B-Instruct2025.07 | 30.2 | |
| ITBase Model=Llama-3-8B-Instruct2026.05 | 30.2 | |
| I-DPOModel=Llama-3-8B-Instruct2025.07 | 29.8 | |
| I-DPOBase Model=Llama-3-8B-Instruct, DPO Mode=iterative2026.05 | 29.8 | |
| DPO + MaPPOModel=Llama-3-8B-Instruct2025.07 | 29.5 | |
| DPO + MaPPOBase Model=Llama-3-8B-Instruct, DPO Mode=offline2026.05 | 29.5 | |
| I-DPO + MaPPOModel=Llama-3-8B-Instruct2025.07 | 28.8 | |
| I-DPO + MaPPOBase Model=Llama-3-8B-Instruct, DPO Mode=iterative2026.05 | 28.8 | |
| DPOModel=Llama-3-8B-Instruct2025.07 | 27.5 | |
| DPOBase Model=Llama-3-8B-Instruct, DPO Mode=offline2026.05 | 27.5 |