JudgeSense: A Benchmark for Prompt Sensitivity in LLM-as-a-Judge Systems
Factual evidence
What the source reports
Research introduces JudgeSense, a benchmark measuring LLM-as-a-judge system sensitivity to semantically equivalent prompt paraphrases, via a Judge Sensitivity Score (JSS).
Inspect the evidence
- Inclusion basis
- Enterprise AI
- Publisher and source type
- arXiv cs.CL — Computation and Language · RESEARCH
- Published by source
- 21 September 2026
- Collected by OneBench
- 28 Apr 2026, 21:40 UK
Stored source excerpt
arXiv:2604.23478v1 Announce Type: new Abstract: Large language models are increasingly deployed as automated judges for evaluating other models, yet the stability of their verdicts under…
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
LLM-as-a-judge systems, currently used for internal model evaluation, face a new validation challenge if prompt sensitivity leads to inconsistent verdicts that undermine model risk and governance frameworks.
Do what
This research identifies a critical vulnerability in current LLM evaluation methods that your model risk team needs to consider when approving internal LLM-as-a-judge deployments.