mlsanityData to 7 Oct 2026

Dissociation of Faithful and Unfaithful Reasoning in LLMs

arXiv 2405.15092 ↗First posted 23 May 2024cs.AI cs.CLcited by 23

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

Abstract (arXiv)

Large language models (LLMs) often improve their performance in downstream tasks when they generate Chain of Thought reasoning text before producing an answer. We investigate how LLMs recover from errors in Chain of Thought. Through analysis of error recovery behaviors, we find evidence for unfaithfulness in Chain of Thought, which occurs when models arrive at the correct answer despite invalid reasoning text. We identify factors that shift LLM recovery behavior: LLMs recover more frequently from obvious errors and in contexts that provide more evidence for the correct answer. Critically, these factors have divergent effects on faithful and unfaithful recoveries. Our results indicate that there are distinct mechanisms driving faithful and unfaithful error recoveries. Selective targeting of these mechanisms may be able to drive down the rate of unfaithful reasoning and improve model interpretability.

Key facts. Cited by 23 papers in the map (data through 7 Oct 2026). Placed in Prompting and chain-of-thought reasoning, one of 17 directions in LLM reasoning. Builds on 1 of its 38 references that are in the map. First posted on arXiv 23 May 2024.

Builds on

1 of 38 references.

Closest ancestors · 1

Same direction, newest first

  1. 2023On Measuring Faithfulness or Self-consistency of Natural Language Explanations13 Nov

Built on by

Cited by year (2026 to 7 Oct)

Citing papers per year: 2024: 1; 2025: 12; 2026: 10.

23 papers.

Sort:

By area:

  1. 2025Chain-of-Thought Reasoning In The Wild Is Not Always Faithfulcited by 166
  2. 2025FaithCoT-Bench: Benchmarking Instance-Level Faithfulness of Chain-of-Thought Reasoningcited by 25
  3. 2025Measuring the Faithfulness of Thinking Drafts in Large Reasoning Modelscited by 22
  4. 2025A Comprehensive Survey on Trustworthiness in Reasoning with Large Language Modelscited by 22
  5. 2025How does Chain of Thought Think? Mechanistic Interpretability of Chain-of-Thought Reasoning with Sparse Autoencodingcited by 19
  6. 2026Learning to Self-Verify Makes Language Models Better Reasonerscited by 18
  7. 2025Improving Chain-of-Thought Reasoning via Quasi-Symbolic Abstractionscited by 17
  8. 2024CoMAT: Chain of Mathematically Annotated Thought Improves Mathematical Reasoningcited by 16
  9. 2025Can Aha Moments Be Fake? Towards Quantifying Decorative and True Thinking in Chain-of-Thoughtcited by 14
  10. 2025State over Tokens: Characterizing the Role of Reasoning Tokenscited by 5
  11. 2026ReasoningFlow: Discourse Structures for Understanding LLM Reasoning Tracescited by 5
  12. 2025Learning to Ponder: Adaptive Reasoning in Latent Spacecited by 4
11 more
  1. 2025Unspoken Hints: Accuracy Without Acknowledgement in LLM Reasoningcited by 4
  2. 2026Measuring Faithfulness Depends on How You Measure: Classifier Sensitivity in LLM Chain-of-Thought Evaluationcited by 4
  3. 2026Detecting and Suppressing Reward Hacking with Gradient Fingerprintscited by 4
  4. 2025Correlation or Causation: Analyzing the Causal Structures of LLM and LRM Reasoning Processcited by 3
  5. 2026Detecting Unfaithful Chain-of-Thought via Circuit-Guided Internal-External Discrepancycited by 3
  6. 2026Bypassing the Rationale: Causal Auditing of Implicit Reasoning in Language Modelscited by 2
  7. 2026Deep Reasoning in General Purpose Agents via Structured Meta-Cognitioncited by 2
  8. 2025Do Biased Models Have Biased Thoughts?cited by 1
  9. 2026Explanation-Bound Tool Execution for AI Agents: Server-Verified Action Claims Without Trusting Model Rationalescited by 1
  10. 2026Forecasting Future Behavior as a Learning Taskcited by 0
  11. 2026Beyond Local Accuracy: A Protocol-Level Identifiability Audit for Controlled LLM Reasoning Evaluationcited by 0

Similar, not cited

Close in wording, no citation either way

  1. 2023Measuring Faithfulness in Chain-of-Thought Reasoning
  2. 2026Fragile Thoughts: How Large Language Models Handle Chain-of-Thought Perturbations
  3. 2026Does Verbose Chain-of-Thought Really Help? In-Distribution Evidence that Content, Not Length, Matters
  4. 2026Legibility is Not Interpretability: Comparing Judged and Actual Importance in Chain-Of-Thought Reasoning
  5. 2025Lost at the Beginning of Reasoning
  6. 2025Process or Result? Manipulated Ending Tokens Can Mislead Reasoning LLMs to Ignore the Correct Reasoning Steps
  7. 2026Reasoning-Token Spikes Under Prompted Untruthful Responding in Large Language Models
  8. 2026LLM Reasoning Is Latent, Not the Chain of Thought