SOURCE-LINKED INTELLIGENCE

EDCT-Bench: Uncovering Faithfulness Gaps in VLMs via Explanation-Driven Counterfactual Testing

arXiv · AI, language, vision and robotics · article · Sep 16, 2026 · UTC

Vision-Language Models (VLMs) can produce Natural Language Explanations (NLEs) that sound plausible yet remain inconsistent with the visual evidence they cite. We present Explanation-Driven Counterfactual Testing (EDCT), an intervention-based protocol that extracts visual concepts cited in a model's explanation, applies verified minimal edits to them, and tests whether the resulting answer and explanation remain consistent with the edited image. Using this protocol, we create EDCT-Bench, a comprehensive benchmark spanning three complementary domains: knowledge-intensive visual question answeri

Read original source ↗ Open in workspace

recordType

paper

region

Global