评分标准作为攻击面：LLM 评判器中的隐蔽偏好漂移

Computer Science > Cryptography and Security arXiv:2602.13576 (cs) [Submitted on 14 Feb 2026 (v1), last revised 14 Sep 2026 (this version, v2)] Title:Rubrics as an Attack Surface: Stealthy Preference Drift in LLM Judges Authors:Ruomeng Ding, Yifei Pang, He Sun, Yizhong Wang, Zhiwei Steven Wu, Zhun Deng View a PDF of the paper titled Rubrics as an Attack Surface: Stealthy Preference Drift in LLM Judges, by Ruomeng Ding and 5 other authors View PDF HTML (experimental) Abstract:Evaluation and alignment pipelines for large language models increasingly rely on LLM-based judges, whose behavior is guided by natural-language rubrics and validated on benchmarks. We identify a previously under-recognized vulnerability in this workflow, which we term Rubric-Induced Preference Drift (RIPD). Even when rubric edits pass benchmark validation, they can still produce systematic and directional shifts in a judge's preferences on target domains. Because rubrics serve as a high-level decision interface, such drift can emerge from seemingly natural, criterion-preserving edits and remain difficult to detect through aggregate benchmark metrics or limited spot-checking. We further show this vulnerability can be exploited through rubric-based preference attacks, in which benchmark-compliant rubric edits steer judgments away from a fixed human or trusted reference on target domains, systematically inducing RIPD and reducing target-domain accuracy up to 9.5% (helpfulness) and 27.9% (harmlessness). When these judgments are used to generate preference labels for downstream post-training, the induced bias propagates through alignment pipelines and becomes internalized in trained policies. This leads to persistent and systematic drift in model behavior. Overall, our findings highlight evaluation rubrics as a sensitive and manipulable control interface, revealing a system-level alignment risk that extends beyond evaluator reliability alone. The code is available at: this https URL. Warning: Certain sections may contain potentially harmful content that may not be appropriate for all readers. Comments: Subjects: Cryptography and Security (cs.CR); Artificial Intelligence (cs.AI); Computation and Language (cs.CL) Cite as: arXiv:2602.13576 [cs.CR] (or arXiv:2602.13576v2 [cs.CR] for this version) https://doi.org/10.48550/arXiv.2602.13576 Focus to learn more arXiv-issued DOI via DataCite Submission history From: Ruomeng Ding [view email] [v1] Sat, 14 Feb 2026 03:19:14 UTC (571 KB) [v2] Mon, 14 Sep 2026 12:28:07 UTC (602 KB) Full-text links: Access Paper: View a PDF of the paper titled Rubrics as an Attack Surface: Stealthy Preference Drift in LLM Judges, by Ruomeng Ding and 5 other authors View PDF HTML (experimental) TeX Source view license Current browse context: cs.CR new | recent | 2026-02 Change to browse by: cs cs.AI cs.CL References & Citations NASA ADS Google Scholar Semantic Scholar export BibTeX citation Loading... BibTeX formatted citation × loading... Data provided by: Bookmark Bibliographic Tools Bibliographic and Citation Tools Bibliographic Explorer Toggle Bibliographic Explorer (What is the Explorer?) Connected Papers Toggle Connected Papers (What is Connected Papers?) Litmaps Toggle Litmaps (What is Litmaps?) scite.ai Toggle scite Smart Citations (What are Smart Citations?) Code, Data, Media Code, Data and Media Associated with this Article alphaXiv Toggle alphaXiv (What is alphaXiv?) Links to Code Toggle CatalyzeX Code Finder for Papers (What is CatalyzeX?) DagsHub Toggle DagsHub (What is DagsHub?) GotitPub Toggle Gotit.pub (What is GotitPub?) Huggingface Toggle Hugging Face (What is Huggingface?) ScienceCast Toggle ScienceCast (What is ScienceCast?) Demos Demos Replicate Toggle Replicate (What is Replicate?) Spaces Toggle Hugging Face Spaces (What is Spaces?) Spaces Toggle TXYZ.AI (What is TXYZ.AI?) Related Papers Recommenders and Search Tools Link to Influence Flower Influence Flower (What are Influence Flowers?) Core recommender toggle CORE Recommender (What is CORE?) Author Venue Institution Topic About arXivLabs arXivLabs: experimental projects with community collaborators arXivLabs is a framework that allows collaborators to develop and share new arXiv features directly on our website. Both individuals and organizations that work with arXivLabs have embraced and accepted our values of openness, community, excellence, and user data privacy. arXiv is committed to these values and only works with partners that adhere to them. Have an idea for a project that will add value for arXiv's community? Learn more about arXivLabs. Which authors of this paper are endorsers? | Disable MathJax (What is MathJax?)