Hallucination Rate Assessment on HEAL Probes SceneMod - SynonymSub
20.4CODS-R1-Distil-LLaMA-8B
Evaluation Results
| Method | Links | ||
|---|---|---|---|
| DS-R1-Distil-LLaMA-8BEnv=BEHAVIOR2026.04 | 20.4 | 10.1 | |
| Gemma-2-9b-itEnv=BEHAVIOR2026.04 | 11.4 | 2.8 | |
| Qwen-14B-InstructEnv=BEHAVIOR2026.04 | 3.9 | 10 | |
| Llama-3-8B-InstructEnv=BEHAVIOR2026.04 | 2.2 | 7.4 | |
| Qwen-14B-Instruct-VCDEnv=BEHAVIOR2026.04 | 1 | 5.7 |