Token-Level Hallucination Detection on HDM-Bench
77.03AUROCRAGognizer
Evaluation Results
| Method | Links | |
|---|---|---|
| RAGognizerDetector Category=White-box, Proxy Generator=Qwen3-4B-Instruct-25072026.04 | 77.03 | |
| HallucinationProbesDetector Category=White-box, Probe=LoRA-KL on LLaMA-3.1-8B (λKL = 0.05)2026.04 | 76.38 | |
| HDM-2-3BDetector Category=Black-box, Same-dataset evaluation=true2026.04 | 74.99 | |
| LettuceDetect-LDetector Category=Black-box2026.04 | 69 | |
| PerplexityDetector Category=White-box, Proxy Generator=Qwen3-4B-Instruct-25072026.04 | 59.6 |