Multimodal Hallucination - VH UniHD Resistance: leaderboard
The "VH UniHD 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 - UniHD). Source: huggingface.co. 9 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | MPLUG-Owl2 | 64.85 | |
| 2 | Llama-VID | 60.16 | |
| 3 | InstructBLIP | 56.41 | |
| 4 | LLaVA-v1.5 | 53.96 | |
| 5 | Video-LLaVA | 48.68 | |
| 6 | FAVOR | 45.1 | |
| 7 | Chat-Univi | 43.09 | |
| 8 | Video-Llama | 41.7 | |
| 9 | Valley | 28.25 |
Interactive version: theaggregate.ai/benchmark?slug=multimodal-hallucination-vh-unihd-resistance · How It Works · Data refreshed daily, snapshot 2026-10-20.