Entity · benchmark

RIO-Bench

benchmarkactiveprovisionalrio-bench-b9df59a9·1 events·first seen Jul 3, 2026

Aliases: RIO-Bench

Co-occurring entities

ViT (Vision Transformer)Towards Robustness against Typographic Attack with Training-free Concept Localization CLIP

More like this (12)

RoleBench RepoBench AARRI-Bench RuBench RMCBench RISEBench LiveBench SorryBench ATE-Bench FoldBench SupraBench CursorBench

Recent events (1)

5arXiv · cs.CL·Jul 3, 2026·source ↗

Training-free mechanistic defense against typographic attacks on CLIP-based vision encoders

Researchers propose a training-free method to defend CLIP-based vision encoders against typographic attacks, where irrelevant text embedded in images biases visual representations toward lexical rather than semantic meaning. The approach uses sampling-based mechanistic interpretability to identify specific Vision Transformer attention heads responsible for encoding lexical information, then applies targeted circuit-level interventions to suppress this behavior. Without any retraining, the method outperforms both supervised and training-free baselines on object classification and improves Visual Question Answering accuracy under typographic attack conditions on RIO-Bench across several state-of-the-art LVLMs.

Evaluation and Benchmarking AI Safety Research ViT (Vision Transformer)Towards Robustness against Typographic Attack with Training-free Concept Localization RIO-Bench +2 more