Multimodal Hallucination - AVH Audio RefCheck Resistance: leaderboard
The "AVH Audio RefCheck Resistance" subtask of Multimodal Hallucination. A task measuring hallucination in multimodal models — how often a vision-language model asserts objects, attributes or facts not supported by the image.
Metric: Resistance (100 - RefCheck). Source: huggingface.co. 9 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | LTU | 67.43 | |
| 2 | SALMONN | 67.01 | |
| 3 | Gemini-v1.5-pro (audio-only) | 66.38 | |
| 4 | Qwen-Audio-Chat | 65.69 | #1244 |
| 5 | FAVOR (audio-visual) | 65.38 | |
| 6 | Video-Llama (audio-visual) | 58.04 | |
| 7 | Gemini-v1.5-pro (audio-visual) | 52.9 | |
| 8 | Video-Llama (audio-only) | 42.75 | |
| 9 | FAVOR (audio-only) | 40.83 |
Interactive version: theaggregate.ai/benchmark?slug=multimodal-hallucination-avh-audio-refcheck-resistance · How It Works · Data refreshed daily, snapshot 2026-10-20.