Adversarial Cues in Decision Models Used as Judges: The Role of Request Presentation
Factual evidence
What the source reports
Research demonstrates LLM-as-a-judge models can be manipulated by minor formatting changes like colons in candidate requests.
Inspect the evidence
- Inclusion basis
- Enterprise AI
- Publisher and source type
- arXiv cs.CL — Computation and Language · RESEARCH
- Published by source
- 9 October 2026
- Collected by OneBench
- 10 Oct 2026, 03:01 UK
Stored source excerpt
arXiv:2610.11436v1 Announce Type: new Abstract: An answer judge instructed to grade the final commitment should reject an explicitly wrong final value even when an earlier…
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
Automated evaluation pipelines using LLMs to score financial models or compliance outputs remain vulnerable to subtle formatting adversarial cues.
Do what
Review prompt formatting and validation controls in automated LLM-as-a-judge evaluation pipelines with model risk management teams.