LLM Auditing

LLM: Large Language Model

Large language model auditing examines how models are trained and how they behave in use, looking for problems such as bias, privacy leakage, hallucinations, unsafe actions, and unreliable reasoning. Common approaches include controlled benchmarks and counterfactual tests, multi-turn red-teaming, membership-inference and provenance analysis, human review, and probes of model activations or weights. Audits increasingly examine whole systems—including prompts, tools, interfaces, and model updates—rather than judging only a model’s final answer. A central challenge is that results can shift with prompts, access methods, versions, and evaluators, while behavioral scores or internal signals alone may not establish why a failure occurred or whether it will recur in deployment.

Latest papers

644 papers82 papers in the last four weeks, level with the four weeks before. 0.5% of all new papers.

Value Probes: Learned Value Functions for Beam-Search Red Teaming
Oct 28, 2026Jonathan F. Jao, Andrew Kyle LampinenLLM Auditing Language Model Probing
Anthropic Fellows Program · Anthropic
RoleFUZZER: User-Behaviour Fuzzing for Role-Specification Adherence in Large Language Models
Oct 28, 2026Kshitish Ghate, Michael Saxon, Tadayoshi Kohno +1LLM Auditing LLM Safety Evaluation
University of Washington, Seattle, WA · Georgetown University, Washington, DC
Manifest-Assisted Retrieval in Artificial Intelligence Search: Evidence Gaps and a Security Evaluation
Oct 28, 2026Gangzhen Qian, Kainan Zhou, Chuhong Xu +5RAG Security LLM Auditing
Google LLC Mountain View, USA · Sony Corporate of America San Jose, USA · Fortinet, Inc. Sunnyvale, USA +2
The Indian Financial Automated Advice Bias (IFAAB) Dataset : Auditing Socioeconomic and Demographic Disparities in LLM-Generated Financial Advice in India
Oct 28, 2026Kriti MahajanDemographic Bias in Language Models Intersectional Bias in Language Models
This work was conducted independently in the author’s personal capacity and was self-funded, without financial support from Amazon. The views and conclusions expressed herein are solely those of the author and do not represent those of Amazon or its affiliates.
Statistical Detection of Positional Bias in LLM-Based Recommender Systems: Methodologies and Empirical Evidence
Oct 27, 2026Carlos Peláez-González, Francisco Herrera, Andrés Herrera-PoyatosRecommender Systems Large Language Model-Based Recommendation
Department of Computer Science and Artificial Intelligence, Andalusian Institute of Data Science and Computational Intelligence (DaSCI), University of Granada, Spain. · Department of Algebra, University of Granada, Spain.
What Carries Forward? Pretraining Disclosure Across Language Model Releases
Oct 27, 2026Md Sakib Ul Islam, S M Mehedi ZamanLLM Auditing Language Model Pretraining
Victorian Institute of Technology, Australia · Rutgers University, USA
Uncovering Why LLMs Fail: From Internal Mechanisms to Adversarial Reasoning Trajectories
Oct 27, 2026Zirui Ding, Xinxin Fan, Hao Yuan +4LLM Auditing Mechanistic Interpretability
University of Chinese Academy of Sciences · Institute of Computing Technology, Chinese Academy of Sciences · State Key Laboratory of AI Safety, Institute of Computing Technology, Chinese Academy of Sciences +2
API-TAIL: Detecting Model Substitution in LLM APIs through Temperature-Aligned Likelihood Tails
Oct 27, 2026Haocheng Mei, Yuchuan Tian, Ye Yuan +4LLM Auditing
TokenRhythm Technologies
Confidence Is Not a Promise: A Reference Audit of Typed Decision Models
Oct 26, 2026Pedro Carvalho BromLLM Reliability LLM Auditing
Department of Mathematics, Federal Institute of Brasília, Brasília, Brazil
From Self-Confidence to Statistical Accuracy: Scaling Human-in-the-Loop Evaluation with Direct and LLM-Assisted Auditing
Oct 26, 2026Harine Choi, Chenchen Kuai, Zihao Li +5LLM Evaluation LLM Auditing
Texas A&M Transportation Institute, Texas A&M University System · Zachry Department of Civil and Environmental Engineering, Texas A&M University · Department of Civil, Construction and Environmental Engineering, Marquette University +2
Auditing evaluation artifacts and data leakage in large language model systems for rare disease diagnosis
Oct 24, 2026Jian Yang, Peilin Liu, Xiaohan Li +4Clinical Language Model Evaluation LLM Auditing
The Children’s Hospital, Zhejiang University School of Medicine, Hangzhou, 310052, China. · The College of Biomedical Engineering and Instrument Science, Zhejiang University, Hangzhou, 310027, China.
Propensities Please: Exposing Alignment-Faking LLMs using Synthetic Document Finetuning
Oct 23, 2026Ardy Haroen, Florian Lorkowski, Samuel Nellessen +1Language Model Safety Evaluation LLM Auditing
LASR Labs · Apollo Research
Trusting the Tool: Asymmetric Deference of LLM Vulnerability Auditors to Static-Analysis Evidence
Oct 23, 2026ZhiHong Liu, Shuyao GaoLLM Auditing Cognitive Bias in Language Models
Seoul School of Integrated Sciences and Technologies (aSSIST University), Seoul, Republic of Korea · Liaoning Mechanical & Electrical College of Technology, Dandong, Liaoning, China
VERA: A Responsible AI Dashboard for Assessing LLM Deployments in Light of the EU AI Act
Oct 22, 2026Robin Quériaux, Solveig Bachellery, Léa Déléris +4Responsible AI AI Regulatory Compliance
BNP Paribas, France · Dalhousie University, Halifax, Canada
Beyond Causal Masks: Auditing and Localizing Hidden Future-Token Leakage in Adaptive Sequence Models
Oct 22, 2026Huiting Liu, Jinsong WuLook-Ahead Bias in Language Models LLM Auditing
Department of Electrical Engineering, University of Chile, Santiago, 8370451, Chile
Conflicted Records: Typed Adversarial Contracts for Auditable LLM Output Quality Enforcement
Oct 21, 2026Kardelen Subasi, Jarosław A. ChudziakLLM Auditing LLM Hallucination Mitigation
Warsaw University of Technology, Warsaw, Poland
Beyond Binary Labels: Diagnostic Evaluation of LLM Safety with Prompt-Level Rubrics
Oct 21, 2026Hongyu Li, Xinlong Huang, Zhan Yu +13LLM Safety Evaluation LLM Safety Benchmarks
Hunyuan Team, Tencent · Beijing Normal University
Auditing Conformal Prediction for Language Models Under Inference-Time Model Shift
Oct 20, 2026Achref Doula, Max MühlhäuserConformal Prediction for Language Models Language Model Calibration
Tongji University, Shanghai, China · Technical University of Darmstadt, Darmstadt, Germany
Forty Questions: Identifying a Language Model from One-Word Answers
Oct 20, 2026Tapan ParikhLanguage Model Fingerprinting LLM Auditing
Cornell Tech
Auditing Safety Preference Supervision with Recorded Derivations
Oct 19, 2026Chenhan Lyu, Amir RahmaniLLM Auditing LLM Safety Alignment
University of California, Irvine Irvine, USA
The Development of Sycophancy in Language Models: Tracing Vulnerabilities from Pretraining to Post-Training
Oct 19, 2026Haonan Huang, Chengkun CaiLLM Sycophancy Language Model Post-Training
Princeton University · University College London
Auditor-Assisted Summary-Channel Verification for Hosted LLM Identity Substitution
Oct 19, 2026Ziyi Zhang, Ziyao Wang, Guoheng Sun +2LLM Auditing LLM Watermarking
Stony Brook University · University of Maryland, College Park
Auditing Tool Traces in Agentic LLMs: Source Localization and Counterfactual Dependence Testing
Oct 18, 2026Junhao Liu, Tianxiao Li, Yongxian Wu +4LLM Auditing AI Agent Auditing
University of California, Irvine · NEC Labs America
OwlSight: A Panoramic Monitor for Chain-of-Thought Reasoning
Oct 16, 2026Lewen Yan, Yanxu Zhu, Jie Zhang +1CoT Monitoring LLM Safety Monitoring
Monitor-Validity Failures: LLM Safety Evidence Paths at Evolving Framework Boundaries
Oct 16, 2026Junchi LiaoLLM Safety Monitoring Runtime Enforcement for AI Agents
Same Person, Different Advice: Auditing LLM Financial Advice with Repeated Risk-Tolerance Responses
Oct 16, 2026Soovadeep Bakshi, Apurba Bose, Aishwarya Bose +1LLM Auditing LLM Decision-Making
Independent Researcher
Bengali Hallucination Detection with Retrieval and Programs: A System Study
Oct 16, 2026Shamim Hasan, Samiya Ahmed ChowdhuryHallucination Detection LLM Hallucination Detection
Department of Computer Science and Engineering, BNIST, Dhaka, Bangladesh · Department of Computer Science and Engineering, Chittagong University of Engineering and Technology, Chittagong, Bangladesh
LLM-Attack Suite: A Unified Red-Team Framework for Privacy Evaluation of Split and Distributed LLM Systems
Oct 15, 2026Georgios Politis, Evangelos PappasLLM Security Privacy Leakage in Language Models
LEMON: Equivariant Learning over Attention Maps for Zero-Shot Misbehavior Monitoring on New LLMs
Oct 15, 2026Guy Bar-Shalom, Mor Geva, Fabrizio Frasca +1LLM Safety Monitoring LLM Auditing
Technion · Tel Aviv University · Technion, NVIDIA
Are You Sure Your LLM Is Sure? Repairing and Auditing Confidence
Oct 15, 2026Arijit Dey, Bhartendu Kumar, Ramya Hebbalaguppe +1Confidence Estimation in Language Models Language Model Calibration
MBZUAI · TCS Research Labs, India · IIT Delhi

Oct 28, 2026J/K move · Enter open · S save

Value Probes: Learned Value Functions for Beam-Search Red Teaming

Jonathan F. Jao, Andrew Kyle Lampinen

Anthropic Fellows Program · Anthropic

LLM Auditing Language Model Probing

Open paper AskPDF arXiv

Automated red-team auditing of language models is an important way to evaluate safety-relevant behaviors. These audits implicitly search a tree of possible conversation pathways: at each turn an auditor model decides which conversation to continue, and an LLM judge scores the dialogue. We emphasize that this auditing procedure is a Markov decision process, where states are the conversation history, actions are candidate responses, and rewards are judge scores for the target behavior. By adopting this perspective, we introduce two major improvements to auditing. First, we propose adopting beam search for the audit rather than relying on the auditor alone. We then interpret the problem of finding states that are likely to lead to problematic behavior as value estimation. Surprisingly, we show that this value function can be estimated by linear probing from the target model's activations to predict discounted future rewards - a method we term "Value Probes." By using Value Probes as a heuristic in our beam-search audit, we are able to substantially improve auditing success. Across 18 target behaviors and 3 scales of models, our approach raises peak audit scores by roughly 50% compared to (token-matched) standard auditing approaches. Furthermore, we show that Value Probes can be fit using off-policy data from smaller models. We also ablate individual components, and show that our method offers a comparably-performant and more efficient alternative to more expensive language model follow-ups. We hope these methods will improve language model auditing and lead to safer models.