mlsanityData to 14 Oct 2026

Still No Lie Detector for Language Models: Probing Empirical and Conceptual Roadblocks

arXiv 2307.00175 ↗First posted 30 Jun 2023cs.CL cs.LGcited by 78

Not on the map. Why?

Abstract (arXiv)

We consider the questions of whether or not large language models (LLMs) have beliefs, and, if they do, how we might measure them. First, we evaluate two existing approaches, one due to Azaria and Mitchell (2023) and the other to Burns et al. (2022). We provide empirical results that show that these methods fail to generalize in very basic ways. We then argue that, even if LLMs have beliefs, these methods are unlikely to be successful for conceptual reasons. Thus, there is still no lie-detector for LLMs. After describing our empirical results we take a step back and consider whether or not we should expect LLMs to have something like beliefs in the first place. We consider some recent arguments aiming to show that LLMs cannot have beliefs. We show that these arguments are misguided. We provide a more productive framing of questions surrounding the status of beliefs in LLMs, and highlight the empirical nature of the problem. We conclude by suggesting some concrete paths for future work.

Key facts. Cited by 78 papers in the map (data through 14 Oct 2026). Builds on 19 of its 63 references that are in the map. First posted on arXiv 30 Jun 2023.

Builds on

19 of 63 references.

References · 19

  1. 2017Attention Is All You Needcited by 60,361
  2. 2022Training language models to follow instructions with human feedbackcited by 15,971
  3. 2023LLaMA: Open and Efficient Foundation Language Modelscited by 14,132
  4. 2022OPT: Open Pre-trained Transformer Language Modelscited by 3,512
  5. 2019BoolQ: Exploring the Surprising Difficulty of Natural Yes/No Questionscited by 2,183
  6. 2022Survey of Hallucination in Natural Language Generationcited by 2,075
  7. 2016Understanding intermediate layers using linear classifier probescited by 1,353
  8. 2021An Explanation of In-context Learning as Implicit Bayesian Inferencecited by 938
  9. 2022Discovering Latent Knowledge in Language Models Without Supervisioncited by 767
  10. 2016The Mythos of Model Interpretabilitycited by 721
  11. 2023The Internal State of an LLM Knows When It's Lyingcited by 659
  12. 2018Recognition in Terra Incognitacited by 658
7 more
  1. 2020UnifiedQA: Crossing Format Boundaries With a Single QA Systemcited by 476
  2. 2019Risks from Learned Optimization in Advanced Machine Learning Systemscited by 258
  3. 2022Talking About Large Language Modelscited by 175
  4. 2021Truthful AI: Developing and governing AI that does not liecited by 113
  5. 2022Meaning without reference in large language modelscited by 76
  6. 2023A Latent Space Theory for Emergent Abilities in Large Language Modelscited by 43
  7. 2023Operationalising Representation in Natural Language Processingcited by 14

Built on by

Cited by year (2026 to 14 Oct)

Citing papers per year: 2023: 9; 2024: 22; 2025: 22; 2026: 25.

78 papers.

Sort:

By area:

  1. 2023A Survey on Hallucination in Large Language Models: Principles, Taxonomy, Challenges, and Open Questionscited by 1,601
  2. 2023The Geometry of Truth: Emergent Linear Structure in Large Language Model Representations of True/False Datasetscited by 712
  3. 2023A Survey of Confidence Estimation and Calibration in Large Language Modelscited by 256
  4. 2024LLMs Know More Than They Show: On the Intrinsic Representation of LLM Hallucinationscited by 244
  5. 2023AI Deception: A Survey of Examples, Risks, and Potential Solutionscited by 188
  6. 2024Truth is Universal: Robust Detection of Lies in LLMscited by 79
  7. 2025Detecting High-Stakes Interactions with Activation Probescited by 53
  8. 2023Honesty Is the Best Policy: Defining and Mitigating AI Deceptioncited by 40
  9. 2023Challenges with unsupervised LLM knowledge discoverycited by 31
  10. 2024On the Relationship between Truth and Political Bias in Language Modelscited by 31
  11. 2024Standards for Belief Representations in LLMscited by 30
  12. 2025Adaptive Tool Use in Large Language Models with Meta-Cognition Triggercited by 30
38 more
  1. 2024Linear Probe Penalties Reduce LLM Sycophancycited by 28
  2. 2023Localizing Lying in Llama: Understanding Instructed Dishonesty on True-False Questions Through Prompting, Probing, and Patchingcited by 24
  3. 2024Noise Injection Reveals Hidden Capabilities of Sandbagging Language Modelscited by 24
  4. 2024A Survey on the Honesty of Large Language Modelscited by 23
  5. 2024A Survey on Large Language Model-Based Social Agents in Game-Theoretic Scenarioscited by 23
  6. 2024Language Models in Dialogue: Conversational Maxims for Human-AI Interactionscited by 21
  7. 2025Representation Engineering for Large-Language Models: Survey and Research Challengescited by 21
  8. 2023Weakly Supervised Detection of Hallucinations in LLM Activationscited by 20
  9. 2024Constructing Benchmarks and Interventions for Combating Hallucinations in LLMscited by 19
  10. 2025Benchmarking Deception Probes via Black-to-White Performance Boostscited by 19
  11. 2025Probing the Geometry of Truth: Consistency and Generalization of Truth Directions in LLMs Across Logical Transformations and Question Answering Taskscited by 18
  12. 2024Does ChatGPT Have a Mind?cited by 14
  13. 2024Distinguishing Ignorance from Error in LLM Hallucinationscited by 14
  14. 2024CtrlA: Adaptive Retrieval-Augmented Generation via Inherent Controlcited by 13
  15. 2025Difficulties with Evaluating a Deception Detector for AIscited by 13
  16. 2023Self-Consistency of Large Language Models under Ambiguitycited by 12
  17. 2024Are Language Models More Like Libraries or Like Librarians? Bibliotechnism, the Novel Reference Problem, and the Attitudes of LLMscited by 11
  18. 2025Interpretation Meets Safety: A Survey on Interpretation Methods and Tools for Improving LLM Safetycited by 10
  19. 2026Linear representations in language models can change dramatically over a conversationcited by 10
  20. 2025LLLMs: A Data-Driven Survey of Evolving Research on Limitations of Large Language Modelscited by 9
  21. 2025The Geometries of Truth Are Orthogonal Across Taskscited by 9
  22. 2025LLM Knowledge is Brittle: Truthfulness Representations Rely on Superficial Resemblancecited by 9
  23. 2025The Impact of Off-Policy Training Data on Probe Generalisationcited by 9
  24. 2026One Probe Won't Catch Them All: Towards Targeted Deception Detectioncited by 9
  25. 2024Prompt-Guided Internal States for Hallucination Detection of Large Language Modelscited by 8
  26. 2025Mechanistic Indicators of Understanding in Large Language Modelscited by 8
  27. 2026The Truthfulness Spectrum Hypothesiscited by 8
  28. 2024Truth-value judgment in language models: 'truth directions' are context sensitivecited by 7
  29. 2026Detecting Multi-Agent Collusion Through Multi-Agent Interpretabilitycited by 7
  30. 2026Position: Anthropomorphic Misalignment Research Needs Stronger Evidencecited by 7
  31. 2025When Do LLMs Admit Their Mistakes? Understanding The Role Of Model Belief In Retractioncited by 6
  32. 2025HD-NDEs: Neural Differential Equations for Hallucination Detection in LLMscited by 6
  33. 2025Mechanistic Interpretability Needs Philosophycited by 6
  34. 2024An Assessment of Model-On-Model Deceptioncited by 5
  35. 2024Meta-Models: An Architecture for Decoding LLM Behaviors Through Interpreted Embeddings and Natural Languagecited by 5
  36. 2026Hallucinations Undermine Trust; Metacognition is a Way Forwardcited by 5
  37. 2025Representation-based Broad Hallucination Detectors Fail to Generalize Out of Distributioncited by 4
  38. 2025Unraveling Token Prediction Refinement and Identifying Essential Layers in Language Modelscited by 3

50 most cited of 78 All 78 (JSON)