Hallucination Detection on GSM8k (PRR)
0.571PRRSelf-FV
Evaluation Results
| Method | Links | |
|---|---|---|
| Self-FVModel=LLaMA2-7B2026.04 | 0.571 | |
| Function VectorModel=LLaMA2-7B2026.04 | 0.518 | |
| UQ_ICLModel=LLaMA2-13B2026.04 | 0.337 | |
| TADBackbone=Llama-3.1 8B, Setup=out-of-domain2025.05 | 0.323 | |
| RAUQBackbone=Llama-3.1 8B, Setup=out-of-domain2025.05 | 0.323 | |
| MINDBackbone=Llama-3.1 8B, Setup=out-of-domain2025.05 | 0.316 | |
| SheepsBackbone=Llama-3.1 8B, Setup=out-of-domain2025.05 | 0.31 | |
| SATRMD+MSPBackbone=Llama-3.1 8B, Setup=out-of-domain2025.05 | 0.307 | |
| Semantic EntropyModel=LLaMA2-13B2026.04 | 0.293 | |
| Semantic EntropyModel=LLaMA2-7B2026.04 | 0.277 | |
| Self-FVModel=LLaMA2-13B2026.04 | 0.27 | |
| Function VectorModel=LLaMA2-13B2026.04 | 0.261 | |
| LookBackLensBackbone=Llama-3.1 8B, Setup=out-of-domain2025.05 | 0.189 | |
| Total EntropyModel=LLaMA2-13B2026.04 | 0.179 | |
| UQ_ICLModel=LLaMA2-7B2026.04 | 0.159 | |
| SAPLMABackbone=Llama-3.1 8B, Setup=out-of-domain2025.05 | 0.152 | |
| Total EntropyModel=LLaMA2-7B2026.04 | 0.107 | |
| FactoscopeBackbone=Llama-3.1 8B, Setup=out-of-domain2025.05 | -0.04 |