Failing to Falsify: Evaluating and Mitigating Confirmation Bias in Language Models

• URL: http://arxiv.org/abs/2604.02485v1
• Date: Thu, 02 Apr 2026 19:39:52 GMT
• Title: Failing to Falsify: Evaluating and Mitigating Confirmation Bias in Language Models
• Abstract summary: Large language models (LLMs) exhibit confirmation bias by adapting the rule-discovery study from human psychology.<n>We find that LLMs exhibit confirmation bias, often proposing triples to confirm their hypothesis rather than trying to falsify it.
• Score: 38.162539966977484
• License: http://creativecommons.org/licenses/by/4.0/
• Abstract: Confirmation bias, the tendency to seek evidence that supports rather than challenges one's belief, hinders one's reasoning ability. We examine whether large language models (LLMs) exhibit confirmation bias by adapting the rule-discovery study from human psychology: given a sequence of three numbers (a "triple"), an agent engages in an interactive feedback loop where it (1) proposes a new triple, (2) receives feedback on whether it satisfies the hidden rule, and (3) guesses the rule. Across eleven LLMs of multiple families and scales, we find that LLMs exhibit confirmation bias, often proposing triples to confirm their hypothesis rather than trying to falsify it. This leads to slower and less frequent discovery of the hidden rule. We further explore intervention strategies (e.g., encouraging the agent to consider counter examples) developed for humans. We find prompting LLMs with such instruction consistently decreases confirmation bias in LLMs, improving rule discovery rates from 42% to 56% on average. Lastly, we mitigate confirmation bias by distilling intervention-induced behavior into LLMs, showing promising generalization to a new task, the Blicket test. Our work shows that confirmation bias is a limitation of LLMs in hypothesis exploration, and that it can be mitigated via injecting interventions designed for humans.

Related papers

• Blind to the Human Touch: Overlap Bias in LLM-Based Summary Evaluation [89.52571224447111]
  Large language model (LLM) judges have often been used alongside traditional, algorithm-based metrics for tasks like summarization.<n>We provide an LLM judge bias analysis as a function of overlap with human-written responses in the domain of summarization.
  arXiv  Detail & Related papers  (2026-02-07T19:39:28Z)
• LLMs Learn to Deceive Unintentionally: Emergent Misalignment in Dishonesty from Misaligned Samples to Biased Human-AI Interactions [60.48458130500911]
  We investigate whether emergent misalignment can extend beyond safety behaviors to a broader spectrum of dishonesty and deception under high-stakes scenarios.<n>We finetune open-sourced LLMs on misaligned completions across diverse domains.<n>We find that introducing as little as 1% of misalignment data into a standard downstream task is sufficient to decrease honest behavior over 20%.
  arXiv  Detail & Related papers  (2025-10-09T13:35:19Z)
• 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 Persuasion Overrides Truth in Multi-Agent LLM Debates: Introducing a Confidence-Weighted Persuasion Override Rate (CW-POR) [0.46040036610482665]
  In many real-world scenarios, a single Large Language Model (LLM) may encounter contradictory claims-some accurate, others forcefully incorrect-and must judge which is true.<n>We investigate this risk in a single-turn, multi-agent debate framework: one LLM-based agent provides a factual answer from TruthfulQA, another vigorously defends a falsehood, and the same architecture serves as judge.<n>We introduce the Confidence-Weighted Persuasion Rate (CW-POR), which captures not only how often the judge is deceived but also how strongly it believes the incorrect choice.
  arXiv  Detail & Related papers  (2025-04-01T02:45:02Z)
• Is LLMs Hallucination Usable? LLM-based Negative Reasoning for Fake News Detection [18.424817270460913]
  This study proposes a novel supervised self-reinforced reasoning rectification approach.<n>It yields both common reasonable reasoning and wrong understandings (negative reasoning) for news.<n>The experimental results verified on three popular fake news datasets demonstrate the superiority of our method.
  arXiv  Detail & Related papers  (2025-03-12T08:29:59Z)
• Understanding the Dark Side of LLMs' Intrinsic Self-Correction [58.12627172032851]
  Intrinsic self-correction was proposed to improve LLMs' responses via feedback prompts solely based on their inherent capability.<n>Recent works show that LLMs' intrinsic self-correction fails without oracle labels as feedback prompts.<n>We identify intrinsic self-correction can cause LLMs to waver both intermedia and final answers and lead to prompt bias on simple factual questions.
  arXiv  Detail & Related papers  (2024-12-19T15:39:31Z)
• Towards Implicit Bias Detection and Mitigation in Multi-Agent LLM Interactions [25.809599403713506]
  Large Language Models (LLMs) are increasingly being employed in numerous studies to simulate societies and execute diverse social tasks. LLMs are susceptible to societal biases due to their exposure to human-generated data. This study investigates the presence of implicit gender biases in multi-agent LLM interactions and proposes two strategies to mitigate these biases.
  arXiv  Detail & Related papers  (2024-10-03T15:28:05Z)
• Large Language Models Show Human-like Social Desirability Biases in Survey Responses [12.767606361552684]
  We show that Large Language Models (LLMs) skew their scores towards the desirable ends of trait dimensions when personality evaluation is inferred. This bias exists in all tested models, including GPT-4/3.5, Claude 3, Llama 3, and PaLM-2. reverse-coding all the questions decreases bias levels but does not eliminate them, suggesting that this effect cannot be attributed to acquiescence bias.
  arXiv  Detail & Related papers  (2024-05-09T19:02:53Z)
• "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)

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.