A Plausible Explanation Is Enough to Mislead Multimodal Judges
Factual evidence
What the source reports
Researchers introduced JUPE, a benchmark showing plausible textual explanations can mislead multimodal LLM judges on visual tasks.
Inspect the evidence
- Inclusion basis
- Enterprise AI
- Publisher and source type
- arXiv cs.CL — Computation and Language · RESEARCH
- Published by source
- 19 October 2026
- Collected by OneBench
- 20 Oct 2026, 03:01 UK
- Original headline
- A Plausible Explanation Is Enough to Mislead Multimodal Judges ↗
Stored source excerpt
arXiv:2610.22631v1 Announce Type: new Abstract: Multimodal large language models (MLLMs) increasingly serve as judges in training and evaluation pipelines, often assessing answers together with explanations.…
Short excerpt from the collected text, not the full source. Use the source link to read it in context.
The factual summary is a OneBench synthesis, not a quotation or independent verification. Collection time is not publication time. Open the source for its full context; related reporting can share the same underlying announcement.
OneBench interpretation
Institutional assessment
So what
Multimodal models used for automated document inspection or verification can be biased by persuasive textual explanations over visual facts.
Do what
Review evaluation pipelines that use multimodal models to judge document processing or verification workflows for potential explanation bias.