Multimodal Hallucination - AVH Audio CrossCheck-explicit Resistance: leaderboard
The "AVH Audio CrossCheck-explicit 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 - CrossCheck-explicit). Source: huggingface.co. 9 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | Gemini-v1.5-pro (audio-visual) | 71.21 | |
| 2 | Gemini-v1.5-pro (audio-only) | 65.34 | |
| 3 | FAVOR (audio-visual) | 64.8 | |
| 4 | LTU | 62.56 | |
| 5 | Video-Llama (audio-visual) | 58.75 | |
| 6 | SALMONN | 57.79 | |
| 7 | Qwen-Audio-Chat | 56.34 | #1244 |
| 8 | FAVOR (audio-only) | 33.31 | |
| 9 | Video-Llama (audio-only) | 31.95 |
Interactive version: theaggregate.ai/benchmark?slug=multimodal-hallucination-avh-audio-crosscheck-explicit-resistance · How It Works · Data refreshed daily, snapshot 2026-10-20.