Whether LLMs Can Navigate Beliefs and Facts Depends on How You Phrase It

• URL: http://arxiv.org/abs/2608.17809v1
• Date: Tue, 18 Aug 2026 14:06:26 GMT
• Title: Whether LLMs Can Navigate Beliefs and Facts Depends on How You Phrase It
• Abstract summary: We show that large language models (LLMs) exhibit a systemic weakness in acknowledging user beliefs grounded in incorrect information.<n>We find that the size and direction of the weakness depend on the verb used to express the belief, with the accuracy gap between factual and false information ranging from +50% on "I vaguely remember" to -14% on "I seriously doubt"<n>Mechanistically, models attend more to false beliefs they fail to confirm, but suppressing this attention at decoding time recovers accuracy only partially and only in some models.
• Score: 7.192001578491251
• License: http://creativecommons.org/licenses/by/4.0/
• Abstract: Humans naturally form and express beliefs in daily communication, e.g., "I think the answer is 3" or "I suppose that's right." Such beliefs inevitably intertwine with fact and knowledge, making the ability to handle them in tandem desirable for large language models (LLMs), as they are increasingly deployed in user-facing settings. Prior work showed that even capable LLMs exhibit a systemic weakness in acknowledging user beliefs grounded in incorrect information. We extend this evaluation to 10 LLMs across 18 epistemic expressions and find that the size and direction of the weakness depend on the verb used to express the belief, with the accuracy gap between factual and false information ranging from +50% on "I vaguely remember" to -14% on "I seriously doubt". We further show that the phenomenon stems from task confusion: models default to fact-checking the underlying claim, overriding the user's stated belief; chains of thought that explicitly fact-check show lower accuracy on false information than those that do not; and a single instruction can reverse the failure across verb families. Mechanistically, models attend more to false beliefs they fail to confirm, but suppressing this attention at decoding time recovers accuracy only partially and only in some models, calling for future work on intervention methods. Our findings clarify prior results and show how fact-checking, a generally desirable behavior, can interfere with belief tracking in LLMs. Our code is available at https://github.com/ngqm/belief-fact-phrasing.

Related papers

• Evaluating Communicative Belief Updates in Large Language Models via Implicature Recognition and Cancellation [24.054030923829387]
  We evaluate the ability of large language models to recognize unspoken beliefs made through implicatures and to understand their updates through implicature cancellation.<n>We create the first expert-annotated implicature cancellation dataset, ImplicatureX, crowdsourced for human judgements of implicatures and their corresponding cancellations.
  arXiv  Detail & Related papers  (2026-07-27T21:35:22Z)
• The Facade of Truth: Uncovering and Mitigating LLM Susceptibility to Deceptive Evidence [49.94160400740222]
  We introduce MisBelief, a framework that generates misleading evidence via collaborative, multi-round interactions.<n>Using MisBelief, we generate 4,800 instances across three difficulty levels to evaluate 7 representative LLMs.<n>Results indicate that while models are robust to direct misinformation, they are highly sensitive to this refined evidence.<n>We propose Deceptive Intent Shielding (DIS), a governance mechanism that provides an early warning signal by inferring the deceptive intent behind evidence.
  arXiv  Detail & Related papers  (2026-01-09T02:28:00Z)
• Beyond Prompt-Induced Lies: Investigating LLM Deception on Benign Prompts [79.1081247754018]
  Large Language Models (LLMs) are widely deployed in reasoning, planning, and decision-making tasks.<n>We propose a framework based on Contact Searching Questions(CSQ) to quantify the likelihood of deception.
  arXiv  Detail & Related papers  (2025-08-08T14:46:35Z)
• When Do LLMs Admit Their Mistakes? Understanding the Role of Model Belief in Retraction [24.49830646625232]
  We define the behavior of acknowledging errors in previously generated answers as "retraction"<n>We demonstrate that retraction is closely tied to indicators of models' internal belief.<n>Experiments show that internal belief causally influences model retraction.
  arXiv  Detail & Related papers  (2025-05-22T03:16:00Z)
• Walk the Talk? Measuring the Faithfulness of Large Language Model Explanations [0.8949668577519213]
  Large language models (LLMs) are capable of generating plausible explanations of how they arrived at an answer to a question.<n>These explanations can misrepresent the model's "reasoning" process, i.e., they can be unfaithful.<n>We introduce a new approach for measuring the faithfulness of LLM explanations.
  arXiv  Detail & Related papers  (2025-04-19T02:51:20Z)
• Rectifying Belief Space via Unlearning to Harness LLMs' Reasoning [36.74368293113009]
  We propose a method to rectify the belief space by suppressing spurious beliefs while simultaneously enhancing true ones.<n>Our approach first identifies the beliefs that lead to incorrect or correct answers by prompting the model to generate textual explanations.<n>We then apply unlearning to suppress the identified spurious beliefs and enhance the true ones, effectively rectifying the model's belief space.
  arXiv  Detail & Related papers  (2025-02-28T00:57:45Z)
• LACIE: Listener-Aware Finetuning for Confidence Calibration in Large Language Models [69.68379406317682]
  We introduce a listener-aware finetuning method (LACIE) to calibrate implicit and explicit confidence markers. We show that LACIE models the listener, considering not only whether an answer is right, but whether it will be accepted by a listener. We find that training with LACIE results in 47% fewer incorrect answers being accepted while maintaining the same level of acceptance for correct answers.
  arXiv  Detail & Related papers  (2024-05-31T17:16:38Z)
• "I'm Not Sure, But...": Examining the Impact of Large Language Models' Uncertainty Expression on User Reliance and Trust [51.542856739181474]
  We show how different natural language expressions of uncertainty impact participants' reliance, trust, and overall task performance. We find that first-person expressions decrease participants' confidence in the system and tendency to agree with the system's answers, while increasing participants' accuracy. Our findings suggest that using natural language expressions of uncertainty may be an effective approach for reducing overreliance on LLMs, but that the precise language used matters.
  arXiv  Detail & Related papers  (2024-05-01T16:43:55Z)
• The Earth is Flat? Unveiling Factual Errors in Large Language Models [89.94270049334479]
  Large Language Models (LLMs) like ChatGPT are in various applications due to their extensive knowledge from pre-training and fine-tuning. Despite this, they are prone to generating factual and commonsense errors, raising concerns in critical areas like healthcare, journalism, and education. We introduce a novel, automatic testing framework, FactChecker, aimed at uncovering factual inaccuracies in LLMs.
  arXiv  Detail & Related papers  (2024-01-01T14:02:27Z)
• Language Models with Rationality [57.37201135072838]
  Large language models (LLMs) are proficient at question-answering (QA) It is not always clear how (or even if) an answer follows from their latent "beliefs"
  arXiv  Detail & Related papers  (2023-05-23T17:04:25Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.