mlsanityData to 14 Oct 2026

Reasoning Isn't Enough: Examining Truth-Bias and Sycophancy in LLMs

arXiv 2506.21561 ↗First posted 12 Jun 2025cs.CL cs.AIcited by 11

On the map: LLM reasoning → Prompting and chain-of-thought reasoning

Abstract (arXiv)

Despite their widespread use in fact-checking, moderation, and high-stakes decision-making, large language models (LLMs) remain poorly understood as judges of truth. This study presents the largest evaluation to date of LLMs' veracity detection capabilities and the first analysis of these capabilities in reasoning models. We had eight LLMs make 4,800 veracity judgments across several prompts, comparing reasoning and non-reasoning models. We find that rates of truth-bias, or the likelihood to believe a statement is true, regardless of whether it is actually true, are lower in reasoning models than in non-reasoning models, but still higher than human benchmarks. Most concerning, we identify sycophantic tendencies in several advanced models (o4-mini and GPT-4.1 from OpenAI, R1 from DeepSeek), which displayed an asymmetry in detection accuracy, performing well in truth accuracy but poorly in deception accuracy. This suggests that capability advances alone do not resolve fundamental veracity detection challenges in LLMs.

Key facts. Cited by 11 papers in the map (data through 14 Oct 2026). Placed in Prompting and chain-of-thought reasoning, one of 17 directions in LLM reasoning. Builds on 1 of its 25 references that are in the map. First posted on arXiv 12 Jun 2025.

Builds on

1 of 25 references.

References · 1

  1. 2025DeepSeek-R1: Incentivizing Reasoning Capability in LLMs via Reinforcement Learningcited by 6,330

Built on by

Cited by year (2026 to 14 Oct)

Citing papers per year: 2025: 4; 2026: 7.

11 papers.

  1. 2025When Identity Skews Debate: Anonymization for Bias-Reduced Multi-Agent Reasoningcited by 21
  2. 2026What Counts as AI Sycophancy? A Taxonomy and Expert Survey of a Fragmented Constructcited by 16
  3. 2026Reward Modeling for Reinforcement Learning-Based LLM Reasoning: Design, Challenges, and Evaluationcited by 7
  4. 2026Pressure, What Pressure? Sycophancy Disentanglement in Language Models via Reward Decompositioncited by 4
  5. 2025User-Assistant Bias in LLMscited by 3
  6. 2025Getting out of the Big-Muddy: Escalation of Commitment in LLMscited by 1
  7. 2026Distortion Instead of Hallucination: The Effect of Reasoning Under Strict Constraintscited by 1
  8. 2026Evaluating Reasoning Models for Queries with Presuppositionscited by 1
  9. 2025Needle in the Web: A Benchmark for Retrieving Targeted Web Pages in the Wildcited by 0
  10. 2026Status Hierarchies in Language Modelscited by 0
  11. 2026When Helpfulness Becomes Sycophancy: Sycophancy is a Boundary Failure Between Social Alignment and Epistemic Integrity in Large Language Modelscited by 0

Similar, not cited

Close in wording, no citation either way

  1. 2026Diagnosing and Mitigating Sycophancy and Skepticism in LLM Causal Judgment
  2. 2025Towards Evaluting Fake Reasoning Bias in Language Models
  3. 2024A Peek into Token Bias: Large Language Models Are Not Yet Genuine Reasoners
  4. 2025Beyond Prompt-Induced Lies: Investigating LLM Deception on Benign Prompts
  5. 2024LLMs Know More Than They Show: On the Intrinsic Representation of LLM Hallucinations
  6. 2025When Thinking LLMs Lie: Unveiling the Strategic Deception in Representations of Reasoning Models
  7. 2026Beyond the Sycophancy Score: How Task, Model, and Pressure Shape LLM Yielding
  8. 2026Stress-Testing LLM Lie Detectors: Role-Play Failures and Spurious Correlations