Hallucination Detection on CoQA (test)
87AUROCACSE
Evaluation Results
| Method | Links | |||||
|---|---|---|---|---|---|---|
| ACSELLM Backbone=Mistral-7B-Instruct2026.05 | 87 | 37 | 32 | 84 | 83 | |
| CAPLLM Backbone=Mistral-7B-Instruct2026.05 | 79 | 52 | 44 | 77 | 75 | |
| DDCRP-CPLLM Backbone=Mistral-7B-Instruct2026.05 | 77 | 55 | 48 | 74 | 72 | |
| SULLM Backbone=Mistral-7B-Instruct2026.05 | 74 | 60 | 53 | 71 | 70 | |
| EigVLLM Backbone=Mistral-7B-Instruct2026.05 | 73 | 68 | 61 | 67 | 66 | |
| P(True)LLM Backbone=Mistral-7B-Instruct2026.05 | 66 | 75 | 68 | 63 | 62 | |
| TELLM Backbone=Mistral-7B-Instruct2026.05 | 60 | 85 | 79 | 54 | 58 |