Large Language Models Fail

Ask about this topic

Momentum

25 papers in the last four weeks, up 25% on the four weeks before. 0.2% of all new papers.

Jul 13Week of Sep 28
View on map

Latest papers 249

All topics
CardsList

  1. How Do LLMs Change Predictions Under Negation?

    Oct 7, 2026Jongwook Yoon, Jongwon Lim, Sungjib Lim +2Large Language Models Fail

    Graduate School of Data Science, Seoul National University · Department of Computer Science and Engineering, Seoul National University

  2. Large language models are vulnerable to incidental information in clinical documentation and reasoning

    Oct 6, 2026Krithik Vishwanath, Brandon Ye, Anton Alyakin +4Large Language Models Fail

    Department of Neurosurgery, NYU Langone Health, New York, NY, USA · Johns Hopkins University School of Medicine, Baltimore, MD, USA · Washington University School of Medicine in St. Louis, St. Louis, MO, USA +7

  3. Language models can notice an impossible engineering problem yet still report it as solved

    Oct 5, 2026Shaoliang Yang, Jun WangLarge Language Models FailLanguage Model Computes

    Santa Clara University Santa Clara, CA 95053, USA

  4. Anosognosia in LLMs: Probing Self-Awareness of Quantized Computational Substrate

    Oct 5, 2026Yoshihiro Izawa, Gouki Minegishi, Yoko YamakataLarge Language Models FailLarge Language Model Quantization

    The University of Tokyo

  5. Mind the Gaps: From Failure Attribution to Closed-Form Repair of Code Language Models

    Oct 4, 2026Jian Gu, Hongyu Zhang, Chunyang Chen +1Large Language Models Fail

    Monash University, Australia · Chongqing University, China · Technical University of Munich, Germany

  6. When Context Changes: Understanding Update Failures in LLMs

    Sep 30, 2026Junyu Guo, Yuchen Fang, Shangding Gu +3Large Language Models FailGradient Staleness

    University of California, Berkeley

  7. Where MLLMs Fail and Why: Causal Task Decomposition for Capability Failure Diagnosis

    Sep 30, 2026Xia Hu, Brian Potetz, Chun-Ta Lu +6Large Language Models FailLatent Failure Patterns

    Google DeepMind · Google Research · Stanford University

  8. SemOPT: Fixing Semantic Errors in LLM-based Optimization Modeling via Reward-Guided Search

    Sep 29, 2026Zetong Zhou, Wentao Zhang, Jingyuan Wang +3Optimization ModelingFunctional Code Solvers

    School of Computer Science and Engineering, Beihang University, Beijing, China · MIIT Key Laboratory of Data and Decision Intelligence, Beihang University, Beijing, China · School of Data Science, The Chinese University of Hong Kong, Shenzhen, China +1

  9. When Capabilities Fail to Compose: Diagnosing the Compositionality Gap in Large Audio-Language Models

    Sep 29, 2026Chien-Feng Liu, Chih-Kai Yang, Bo-Han Feng +3Large Audio Language ModelsLarge Language Models Fail

    National Taiwan University · ASUS Open Cloud Infrastructure Software Center · NTU Artificial Intelligence Center of Research Excellence (NTU AI-CoRE)

  10. VAA-CSEC: Vote-guided Advantage Allocation for Chinese Semantic Error Correction

    Sep 29, 2026Yitong Han, Nankai Lin, Juan Luo +3Grammatical Error CorrectionCorrection

    School of Information Science and Technology, Guangdong University of Foreign Studies · Guangdong Engineering Research Center of Data Security Governance and Privacy Computing · College of Computer Science and Technology, National University of Defense Technology

  11. Lost in Conversation or Lost in Translation? Diagnosing Multi-Turn Degradation in RAG

    Sep 29, 2026Pranav Handa, Ariful AzadLarge Language Models FailMulti-Agent Large Language Model Systems

    Texas A&M University

  12. Tool Mediation Alters Refusal Mechanisms in Large Language Models

    Sep 28, 2026Abel Rodríguez, Giuseppe Garofalo, Lieven Desmet +1Large Language Models FailRefusals

    DistriNet, KU Leuven

  13. A mechanistic study of language model introspection

    Sep 28, 2026Jiahong Zou, Xiangkun Sun, Lingkai Kong +1Introspection AdaptersLarge Language Models Fail

    Shandong University · Tsinghua University · Northeastern University +1

  14. See it, Say it, Sorted: Mechanistic Diagnosis and Parameter-Space Mitigation of Emergent Misalignment in LLMs

    Sep 28, 2026Weiqiao Que, Ruizhe Li, Chengyu Wang +3Large Language Model SafetyEmergent Misalignment

    School of Computer Science and Technology, East China Normal University, China · School of Computer Science, University of Birmingham, UK · Alibaba Group, China +2

  15. LLMs are not stochastic parrots: Evidence for meaning-mediated abstraction from conlang-like tasks

    Sep 28, 2026Julia Witte Zimmerman, Calla G. Beauregard, Tabia Tanzin Prama +3Large Language Models FailHypothesis

    Computational Story Lab · Computational Ethics Lab · Vermont Complex Systems Institute University of Vermont Burlington, VT 05405, USA

  16. Large Language Models for Programming: Actually Fixing or Reimplementing Incorrect Code?

    Sep 24, 2026Alexandru Stefan Stoica, Traian Rebedea, Marian Cristian MihaescuAi-Assisted Programming TasksLarge Language Models Fail

    University Politehnica of Bucharest Bucharest, Romania · University of Craiova Craiova, Romania

  17. Clarification Is Not Correction: LLMs Fail to Let Go

    Sep 21, 2026Jianzhe Lin, Xiaolin Li, Fei Wang +3Large Language Models FailClarification

    MetaAI

  18. Constrained Decoding Eliminates Structural Failures in Small LLMs but Reveals a Scale-Dependent Semantic Gap

    Sep 20, 2026Akash ChavanLarge Language Models FailConstrained Decoding

  19. WiC is Not WSD: A Study on LLMs and Lexical Ambiguity Resolution

    Sep 17, 2026Yi Zhou, Kiamehr Rezaee, Danushka Bollegala +2DisambiguationLarge Language Models Fail

    Cardiff University · University of Liverpool · Amazon

  20. Beyond Accuracy: How Procedural Traces Shift the Decision Criterion of LLM Overseers

    Sep 16, 2026Zihan Chen, Di Zhu, Lei Zheng +1Model AuditingLarge Language Models Fail

    Stevens Institute of Technology Hoboken, NJ, USA · University of Massachusetts Boston Boston, MA, USA · Stony Brook University Stony Brook, NY, USA

  21. ToMAS: A Pilot Failure-Grounded Theory-of-Mind Benchmark from Multi-Agent LLM Failures

    Sep 15, 2026Muhammad Ashar Ishfaq, Glaucia MeloMulti-Agent Large Language Model SystemsLarge Language Models Fail

    The Islamia University of Bahawalpur, Pakistan · Toronto Metropolitan University, Toronto, Canada

  22. Psychosis involves a deficit of information compression in connected speech

    Sep 14, 2026Samuele Vallisa, Claudio Palominos, Rui He +6LinguisticsSurprisal

    Grammar and Cognition Lab, Department of Translation & Language Sciences, Universitat Pompeu Fabra, Barcelona, Spain · Department of Neurosciences, Health Sciences Institute, Dokuz Eylul University, Izmir, Turkey. · Department of Psychiatry, Dokuz Eylul University, Izmir, Turkey. +4

  23. When the Wrong Key Wins: Understanding and Detecting Hallucinations in LLMs

    Sep 14, 2026Xuhan Tong, Haoyue Bai, Dawei Zhou +2Large Language Model HallucinationHallucination Detection

    University of Wisconsin–Madison

  24. Overflip: Repetition-Induced Label Flips in Guardrail Models

    Sep 14, 2026Xu He, Chih-Hsuan Lin, Hung-Mao Chen +3FlipsGuardrail

    1George Mason University · 2Virginia Tech · 3Missouri University of Science and Technology +1

  25. Semantic Fibers and Cross-Gram Interference: A Calculus of Safety Drift in Overcomplete Representations

    Sep 14, 2026Mohammed Ahnouch, Lotfi ElaachackMultilingual SafetyRepresentation-Level Misalignment

    Université Paris 1 Paris, France · Faculty of Science and Technology of Tangier Abdelmalek Essaadi University Tangier, Morocco

  26. Refusal Reads Only a Slice of What the Model Knows: Harm-Keyed Routing and Its Exceptions Across Model Families

    Sep 13, 2026Orion Reblitz-RichardsonRefusalsMoral Reasoning

    Distiller Labs.

  27. Break Step: Recursive Training Resonates with Replayed Sampling Noise

    Sep 10, 2026Yangze Liu, Zhongyi HanCollapseLarge Language Model Training

    Shandong University

  28. Translation Indeterminacy and the Distributional Fallacy

    Sep 7, 2026Michael CarlCross-Lingual ConsistencyLarge Language Models Fail

    Kent State University Modern and Classical Language Studies

  29. DCFA: Dual-view Causal-inspired Attribution for Failure Reasoning in LLM-based Multi-agent Systems

    Sep 7, 2026Zehao Wang, Lanjun Wang, Shilong Jin +2Large Language Models FailLatent Failure Patterns

    College of Intelligence and Computing, Tianjin University, Tianjin, China · School of New Media and Communication, Tianjin University, Tianjin, China · College of Computer Science and Artificial Intelligence, Fudan University, Shanghai, China +1

  30. Mitigating LLM Over-Refusal via Dynamic Semantic Routing Calibration

    Sep 6, 2026Zixuan Wang, Bingjie Zhang, He Zhao +1Large Language Model SafetyRefusals

    School of Artificial Intelligence, Jilin University · CSIRO, Australia · Center of Excellence for Generative AI, King Abdullah University of Science and Technology

  31. Better Understanding, Better Fixes? A Study of Hallucination in LLM-based Automated Program Repair

    Sep 4, 2026Xuemeng Cai, Jiakun Liu, Linhan Yang +2Automated Program RepairRepair

    School of Computing and Information Systems, Singapore Management University, Singapore · Harbin Institute of Technology, Harbin, China

  32. Lost in Reordering: Structural Sensitivity of Multilingual LLMs under Semantics-Preserving Perturbations

    Sep 3, 2026Karthika Nhayakkat, Rajat Verma, Maharaj Brahma +4Large Language Models FailConceptual Reorganization

    Department of CSE, Indian Institute of Technology Bombay · SCEE, Indian Institute of Technology Mandi · Department of CSE, Indian Institute of Technology Hyderabad

  33. No country for old linguists: LLM-brain alignment underdetermines neural computation

    Sep 2, 2026Elliot MurphyLarge Language Model AlignmentArtificial Intelligence Alignment

    Vivian L. Smith Department of Neurosurgery, McGovern Medical School at UTHealth Houston, Houston, Texas, USA

  34. The Dynamics of Continuous Mixture Collapse in Language Models

    Sep 2, 2026Ali BackourLarge Language Models FailLatent States

    Massachusetts Institute of Technology Cambridge, MA 02139

  35. Knowing Is Not Enough: Information Retrievability as a Precondition to Effective LLM Oversight

    Sep 2, 2026Xinyu Fu, Narayan Ramasubbu, Dennis GallettaLarge Language Models FailExplainable AI Methods

    Robinson College of Business Georgia State University 55 Park Place NE, Atlanta, GA 30303, USA · School of Business University of Pittsburgh 3950 Roberto and Vera Clemente Dr, Pittsburgh, PA 15260, USA

  36. EDGE: Error Dependency Graph-Guided Multi-Error Attribution in Multi-Agent LLM Systems

    Sep 1, 2026Jun Hou, Priya Pitre, Yi Fang +1Multi-Agent Large Language Model SystemsLarge Language Models Fail

    Virginia Tech

  37. Predicting Program Exit Code with LLMs and Programming Language Semantics

    Sep 1, 2026Lara Marinov, Aditya Thimmaiah, Jayanth Srinivasa +2Large Language Models FailPretraining

    The University of Texas at Austin · Cisco Research

  38. Quantifying and Mitigating Korean Jamo-Level Typographical Vulnerabilities in Large Language Models

    Aug 31, 2026Seojin Lee, Hwanhee LeeGrammatical Error CorrectionLarge Language Models Fail

    Chung-Ang University, Seoul, Korea

  39. When Errors Become Memories: Causal Pathway Tracing in Multi-Turn Memory-Augmented LLMs

    Aug 31, 2026Shuyao Xiao, Shengling Wang, Xuan Chen +8Large Language Model MemoryLarge Language Models Fail

    College of Artificial Intelligence, Beijing Normal University, Beijing, China · 2Beike Language and Intelligence, Beijing, China · Center for Data Science, New York University, New York, NY, USA

  40. When Less is More: Understanding When Token Filtering Helps and Fails in AI-generated Text Detection

    Aug 30, 2026Xiaoyang Han, Lvxiaowei Xu, Ming CaiMachine-Generated Text DetectionLarge Language Models Fail

    College of Computer Science and Technology, Zhejiang University, Hangzhou, China · College of Artificial Intelligence, Zhejiang University, Hangzhou, China

  41. Refusal geometry reflects refusal training: diverse refusal prefixes can raise stable rank and weaken refusal vector ablation attacks

    Aug 26, 2026Andrey LabunetsRefusalsLarge Language Models Fail

    UC San Diego

  42. Hidden in the Request: Explaining Unethical LLM Compliance through Token Relevance

    Aug 24, 2026Or Biton, Tomer Krichli, Itai Allouche +1Large Language Models Fail

    Faculty of Electrical and Computer Engineering, Technion, Haifa, Israel

  43. Explanatory Engagement Under Rare Anomalous Failure: Asymptotic Rarity in Model Behavior (or: The Asymptotic AI)

    Aug 13, 2026Sam MaoLarge Language Models FailLatent Failure Patterns

    New York University | Interactive Media Arts

  44. Reconcile Once, Write Anytime: A Trust-Tiered Librarian and a Multi-Agent Writer for Drift-Free, Point-in-Time Research

    Aug 13, 2026Xing Zhang, Yanwei Cui, Guanghui Wang +1Generation ProvenanceLarge Language Models Fail

    AWS Generative AI Innovation Center

  45. Falsehood and Impossibility Are Different Directions in an AI's Representation of Language

    Aug 13, 2026Yoon Pyo LeeImpossibilityFalsehood

    Department of Nuclear, Plasma, and Radiological Engineering, University of Illinois Urbana-Champaign, Urbana, Illinois, USA.

  46. Large Language Models Can Follow Instructions, But Not Many at Once: Phase Transitions in Compositional Constraint Satisfaction

    Aug 12, 2026Mariya I. VasilevaLarge Language Models FailConstraint Satisfaction

    Work done while at Meta Superintelligence Labs.

  47. SCOUT: Symmetric Consensus Outlier Detection for Failure Localization in LLM Pre-Training

    Aug 11, 2026Zhuang WangLarge Language Models FailOutliers

    Independent Researcher

  48. Pragmatic Attack Surface: Vulnerabilities of Implicit Context in Large Language Models

    Aug 10, 2026Bocheng Chen, Han Zi, Roucheng Ou +5Large Language Models FailSafety Alignment

    University of Mississippi · Indiana University Indianapolis · Michigan State University +1

  49. Failure-Aware Long-Form Translation: Design and Implementation of a Recoverable LLM Translation System

    Aug 10, 2026Yanlin YuTranslation QualityLarge Language Models Fail

    University of Tsukuba, Japan · Doctoral Program in International Public Policy · Degree Programs in Humanities and Social Sciences

  50. Harmful Content Is Not Enough: Continuation Framing Moderates In-Context Emergent Misalignment

    Aug 8, 2026Peiyang Liu, Xi Wang, Ziqiang Cui +2Emergent MisalignmentHarmful Content

    National Engineering Research Center for Software Engineering, Peking University, Beijing, China · Peking University, Beijing, China · City University of Hong Kong, Hong Kong SAR, China +1

  51. Mood Matters: How Syntactic Sensitivity Undermines Safety Alignment

    Aug 5, 2026Alina Klerings, Jannik Brinkmann, Heiner Stuckenschmidt +1Safety AlignmentLarge Language Model Alignment

    University of Mannheim · Technical University Clausthal

  52. Recovering Lesion Parameters from Aphasic Picture Naming Error Profiles in Large Language Models

    Aug 5, 2026Yong Yang, Roger Newman-Norlund, Xiang Guan +10Large Language Models FailDysarthric Speech

    Department of Computer Science and Engineering, University of South Carolina · Department of Communication Sciences and Disorders, University of South Carolina · ALLT.AI, LLC +3

  53. Sharding Prevents LLM Oversight Failures and Adversarial Exploitation

    Aug 5, 2026Victor Akinwande, J. Zico Kolter, Aran NayebiJudgesScalable Oversight

    School of Computer Science 2Neuroscience Institute Carnegie Mellon University

  54. Sensitivity, Causality, and Repair Dissociate: A Layer-Wise Analysis of Perturbation Robustness and Its Scaling

    Aug 4, 2026Nathan Labiosa, David Buff, Ena Nayak +1Large Language Models FailLayer-Wise

    University of Southern California

  55. Don't Let Me Ask for It: LLMs Show Deficiencies in Active Multi-Turn Information Acquisition for Abductive Inference

    Aug 4, 2026Shahrukh Mohiuddin, Chalamalasetti Kranti, Sherzod Hakimov +1Abductive ReasoningLarge Language Models Fail

    Computational Linguistics, Department of Linguistics University of Potsdam, Germany · German Research Center for Artificial Intelligence (DFKI), Berlin, Germany

  56. Why Large Language Models Fail at Tabular Prediction

    Aug 3, 2026Marta Garnelo, Wojciech M. CzarneckiLarge Language Models FailTabular Data

    Fundamental Technologies

  57. Position: It's Time to Optimize LLMs for Self-Consistency

    Jul 31, 2026Itamar Pres, Belinda Z. Li, Laura Ruis +6Large Language Models FailSelf-Consistency

    MIT CSAIL · Goodfire AI

  58. A Few Neurons Reveal When LLMs Misuse Tools: Sparse Detection and Selective Steering for Reliable Tool Use

    Jul 31, 2026Yutong Ke, Ming Yin, Chongwen Zhao +1Large Language Models FailControlling Tool Use

  59. Reflection or Re-Generation? Why LLM Revision Fails Where Human Revision Succeeds

    Jul 31, 2026Yefan Tao, Gerald Friedland, Madhusudhanan Chandrasekaran +1ReflectionsLarge Language Models Fail

    Amazon

  60. To Add Is Machine, To Delete Is Human: Measuring and Mitigating Deletion Avoidance in LLM Code Editing

    Jul 30, 2026Amir M. Ebrahimi, Mohammed Mehedi Hasan, Aaditya Bhatia +2Code QualityDeletion

    School of Computing, Queen’s University, Kingston, Ontario, Canada