turing-motors/RIO-Bench

Read or Ignore? A Unified Benchmark for Typographic-Attack Robustness and Text Recognition in Vision-Language Models Real-world VLMs must decide when to read text and when to ignore it, e.g., reading traffic signs but not being fooled by text-based attacks on objects. We propose a unified benchmark, RIO-Bench, to evaluate both typographic-attack robustness and text recognition in VLMs through a novel task called RIO-VQA. Problem Settings: VLMs Must Adaptively Read… See the full description on the dataset page: https://huggingface.co/datasets/turing-motors/RIO-Bench.

<iframe src="https://silytics.si/embed/item/c96a3d09-73af-44a1-ac78-4a262beb4fa6" width="420" height="190" style="border:0" loading="lazy" title="Silytics entry"></iframe>

Silytics shows public metadata from Hugging Face and links to the original. Information may be incomplete or out of date. Last refreshed 18 October 2026. Licenses and terms of the original apply to the work itself. Not financial, legal or professional advice.