Hallucination Detection on Summedits (test)
93AUCSeparate Guardrail 9B
Evaluation Results
| Method | Links | |
|---|---|---|
| Separate Guardrail 9BBase model=Gemma2-9B2025.05 | 93 | |
| Separate Guardrail 7BBase model=Mistral-7B-v0.32025.05 | 92 | |
| DSA:LSTBase model=Mistral-7B-v0.32025.05 | 90 | |
| DSA:LSTBase model=Gemma2-9B2025.05 | 89 | |
| DSA:LSTBase model=Qwen2.5-7B2025.05 | 88 | |
| Separate Guardrail 7BBase model=Qwen2.5-7B2025.05 | 88 | |
| DSA:LSTBase model=Qwen2.5-14B2025.05 | 88 | |
| Separate Guardrail 14BBase model=Qwen2.5-14B2025.05 | 88 | |
| DSA:PLRBase model=Qwen2.5-7B2025.05 | 83 | |
| DSA:RTBBase model=Qwen2.5-7B2025.05 | 83 | |
| DSA:RTBBase model=Qwen2.5-14B2025.05 | 83 | |
| DSA:RTBBase model=Gemma2-9B2025.05 | 83 | |
| DSA:PLRBase model=Qwen2.5-14B2025.05 | 81 | |
| Suffix-ZSPromptBase model=Gemma2-9B2025.05 | 78 | |
| DSA:RTBBase model=Mistral-7B-v0.32025.05 | 72 | |
| Suffix-ZSPromptBase model=Qwen2.5-14B2025.05 | 71 | |
| DSA:PLRBase model=Gemma2-9B2025.05 | 70 | |
| Suffix-ZSPromptBase model=Qwen2.5-7B2025.05 | 64 | |
| Separate Guardrail 51MBase model=Qwen2.5-14B2025.05 | 62 | |
| Suffix-ZSPromptBase model=Mistral-7B-v0.32025.05 | 62 | |
| DSA:PLRBase model=Mistral-7B-v0.32025.05 | 62 | |
| Separate Guardrail 25MBase model=Qwen2.5-7B2025.05 | 61 | |
| Separate Guardrail 27MBase model=Mistral-7B-v0.32025.05 | 59 | |
| Separate Guardrail 31MBase model=Gemma2-9B2025.05 | 59 |