Hallucination Detection on HDMBench (test)
85HF1CALAMRFLOW
Evaluation Results
| Method | Links | ||||||
|---|---|---|---|---|---|---|---|
| CALAMRFLOWAveraged over=5 random seeds2026.05 | 85 | 76.2 | 96.1 | 51.2 | 74.6 | 53.4 | |
| CALAMRFLOWseeds=5, test-set size (N)=5202026.05 | 85 | — | — | 51.2 | 74.6 | — | |
| HDM-22026.05 | 73.6 | 74.8 | 74.4 | — | 74.4 | — | |
| HDM-2test-set size (N)=5202026.05 | 73.6 | — | — | — | 74.4 | — | |
| Qwen2026.05 | 60.4 | 49.1 | 78.6 | — | 56.7 | — | |
| Qwentest-set size (N)=5202026.05 | 60.4 | — | — | — | 56.7 | — | |
| GPT-4o2026.05 | 58.7 | 68.4 | 51.4 | — | 69.6 | — | |
| GPT-4otest-set size (N)=5202026.05 | 58.7 | — | — | — | 69.6 | — | |
| GPT-4o-mini2026.05 | 57.7 | 68.4 | 49.9 | — | 69.3 | — | |
| GPT-4o-minitest-set size (N)=5202026.05 | 57.7 | — | — | — | 69.3 | — |