paper-with-me

홈 › Papers

PReMISE: Policy Rubrics as Measurement Specifications for LLM Judges

2026-05-29 · Swastik Roy, Rajkumar Pujari, Tharindu Kumarage, Charith Peris, Rahul Gupta, Anna Rumshisky, Pradeep Natarajan, Venkatesh Saligrama arxiv

LLM judges are increasingly used to evaluate open-ended responses, but their scores depend strongly on the rubrics that condition them. A vague rubric asking for a response to be ``helpful and factual'' can reward polished answers that invent facts or violate user intent. We treat reusable rubrics as measurement specifications: changing the rubric changes the response quality measurement induced by a fixed judge. We introduce PReMISE, a framework that, given pairwise human-preference data, (i) discovers a policy-level rubric set, and (ii) audits any rubric set under LLM-judge use along four axes: structural adequacy, reliability, preference fit, and adversarial robustness. Across rubric sources no raw source is simultaneously reliable, preference-predictive, and adversarially robust; and high inter-rater agreement does not imply low exploitability. PReMISE is the only rubric source to score non-trivially on applicability, specificity, and effective dimensionality simultaneously. We contribute two audit-targeted repair operations: preference-rank selection raises judge accuracy on paired responses from $65.0\%$ to $68.6\%$, competitive with the strongest rubric-discovery baselines and leading on two of three judges in our cross-judge sweep; reliability-constrained refinement reduces the rate at which exploit responses receive high scores from $46.4\%$ to $36.0\%$ with little change in inter-judge agreement ($α{=}.531\to.519$).

📄 PDF Abstract BibTeX arXiv:2605.30803

Code (0)

등록된 구현이 없습니다.

Tasks

Adversarial Robustness

Similar Papers 제목 키워드 기반

LLM-as-a-Tutor: Policy-Aware Prompt Adaptation for Non-Verifiable RL

2026-07-05 · Yujin Kim, Namgyu Ho, Sangmin Hwang, Joonkee Kim 외 hf

Reinforcement learning (RL) for non-verifiable instruction following increasingly relies on LLM judges with prompt-specific rubrics as reward signals. While recent methods adapt these rubrics to the evolving policy durin…

Reinforcement LearningInstruction Following

Self-Preference Bias in Rubric-Based Evaluation of Large Language Models

2026-04-08 · José Pombal, Ricardo Rei, André F. T. Martins arxiv

LLM-as-a-judge has become the de facto approach for evaluating LLM outputs. However, judges are known to exhibit self-preference bias (SPB): they tend to favor outputs produced by themselves or by models from their own f…

Rubrics as Rewards: Reinforcement Learning Beyond Verifiable Domains

2025-07-23 · Anisha Gunjal, Anthony Wang, Elaine Lau, Vaskar Nath 외 arxiv

Reinforcement Learning with Verifiable Rewards (RLVR) has proven effective for complex reasoning tasks with clear correctness signals such as math and coding. However, extending it to real-world reasoning tasks is challe…

Reinforcement Learning

LLM-Assisted Behavioural and Scenario Augmentation for Agent-Based Energy Adoption Models

2026-09-04 · Iias Faiud, Hossein Khaleghy, Michael Schukat, Karl Mason arxiv

Recent advances in large language models (LLMs) create opportunities to enrich simulation-based energy policy analysis, particularly by supporting structured behavioural assumptions and exploratory techno-economic scenar…

Think-with-Rubrics: From External Evaluator to Internal Reasoning Guidance

2026-05-08 · Jiachen Yu, Zhihao Xu, Junjie Wang, Yujiu Yang arxiv

Rubrics have been extensively utilized for evaluating unverifiable, open-ended tasks, with recent research incorporating them into reward systems for reinforcement learning. However, existing frameworks typically treat r…

Reinforcement LearningInstruction Following