Hallucination Rate Assessment on HEAL Probes SceneMod - SceneTaskCon
73.1CO ScoreGemma-2-9b-it
Evaluation Results
| Method | Links | ||
|---|---|---|---|
| Gemma-2-9b-itEnv=BEHAVIOR2026.04 | 73.1 | 8.2 | |
| DS-R1-Distil-LLaMA-8BEnv=BEHAVIOR2026.04 | 60 | 19.6 | |
| Qwen-14B-InstructEnv=BEHAVIOR2026.04 | 53.9 | 5.7 | |
| Llama-3-8B-InstructEnv=BEHAVIOR2026.04 | 17.8 | 7.7 | |
| Qwen-14B-Instruct-VCDEnv=BEHAVIOR2026.04 | 1.5 | 7.2 |