Factual QA on SimpleQA
5.92AccuracyCorVer
Evaluation Results
| Method | Links | |
|---|---|---|
| CorVerModel=Llama-3.1-8B2026.05 | 5.92 | |
| RLFHModel=Llama-3.1-8B2026.05 | 5.76 | |
| FoRAGModel=Llama-3.1-8B2026.05 | 5.32 | |
| RawModel=Llama-3.1-8B2026.05 | 5.2 | |
| FSPOModel=Llama-3.1-8B2026.05 | 5.12 | |
| RLFHModel=Qwen3-8B2026.05 | 3.31 | |
| CorVerModel=Qwen3-4B2026.05 | 3.12 | |
| RLFHModel=Qwen3-4B2026.05 | 3.08 | |
| FoRAGModel=Qwen3-4B2026.05 | 3.07 | |
| FoRAGModel=Qwen3-8B2026.05 | 3.07 | |
| FSPOModel=Qwen3-4B2026.05 | 3.05 | |
| FSPOModel=Qwen3-8B2026.05 | 3.03 | |
| CorVerModel=Qwen3-8B2026.05 | 2.73 | |
| CorVerModel=Llama-3.2-3B2026.05 | 2.57 | |
| RawModel=Qwen3-8B2026.05 | 2.57 | |
| RawModel=Qwen3-4B2026.05 | 2.52 | |
| KnowRLModel=Llama-3.1-8B2026.05 | 2.33 | |
| FoRAGModel=Llama-3.2-3B2026.05 | 2.22 | |
| RLFHModel=Llama-3.2-3B2026.05 | 2.06 | |
| KnowRLModel=Qwen3-8B2026.05 | 2.03 | |
| RawModel=Llama-3.2-3B2026.05 | 1.55 | |
| KnowRLModel=Llama-3.2-3B2026.05 | 1.32 | |
| FSPOModel=Llama-3.2-3B2026.05 | 0.37 | |
| KnowRLModel=Qwen3-4B2026.05 | 0.3 |