| Dataset Name | SOTA Method | Metric | Trend | ||
|---|---|---|---|---|---|
| AMBER | CHAIR_s10.6 | 56 | 2d ago | ||
| HallusionBench | ResDec | Answer Accuracy (aAcc)71.6 | 39 | 1mo ago | |
| Object-HalBench | OmniLMM + RLAIF-V BoN | Mention Hallucination Rate2.6 | 39 | 3mo ago | |
| AMBER (test) | Dyn. CEI | CHAIR5.6 | 38 | 3mo ago | |
| Bingo | LEAD | Bingo Score3.85 | 21 | 1mo ago | |
| MMHalu | LEAD | MMHalu Score4.27 | 21 | 1mo ago | |
| RSHalluEval 1.0 (test) | Qwen2-VL | HF Information Accuracy0.9792 | 21 | 3mo ago | |
| MHumanEval | LLaVA-RLHF | Response Rate72.6 | 20 | 3mo ago | |
| CHAIR | CS52.8 | 12 | 8d ago | ||
| KPM-HA 1.0 (test) | LLaVA-OneVision | GPT-Hall Score2.71 | 11 | 3mo ago | |
| MSCOCO 500 samples (train and val) | InstructBLIP | CHAIR_I12.2 | 6 | 3mo ago | |
| MSCOCO (val) | InstructBLIP | CHAIR Instance Score10.7 | 3 | 3mo ago |