mlsanityData to 7 Oct 2026

FACT-E: Causality-Inspired Evaluation for Trustworthy Chain-of-Thought Reasoning

arXiv 2604.10693 ↗First posted 12 Apr 2026cs.AIcited by 3

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

Abstract (arXiv)

Chain-of-Thought (CoT) prompting has improved LLM reasoning, but models often generate explanations that appear coherent while containing unfaithful intermediate steps. Existing self-evaluation approaches are prone to inherent biases: the model may confidently endorse coherence even when the step-to-step implication is not valid, leading to unreliable faithfulness evaluation. We propose FACT-E, a causality-inspired framework for evaluating CoT quality. FACT-E uses controlled perturbations as an instrumental signal to separate genuine step-to-step dependence from bias-driven artifacts, producing more reliable faithfulness estimates (textitintra-chain faithfulness). To select trustworthy trajectories, FACT-E jointly considers textitintra-chain faithfulness and textitCoT-to-answer consistency, ensuring that selected chains are both faithful internally and supportive of the correct final answer. Experiments on GSM8K, MATH, and CommonsenseQA show that FACT-E improves reasoning-trajectory selection and yields stronger in-context learning exemplars. FACT-E also reliably detects flawed reasoning under noisy conditions, providing a robust metric for trustworthy LLM reasoning.

Key facts. Cited by 3 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 29 of its 56 references that are in the map. First posted on arXiv 12 Apr 2026.

Builds on

29 of 56 references.

Closest ancestors · 11

Same direction, newest first

  1. 2025FaithCoT-Bench: Benchmarking Instance-Level Faithfulness of Chain-of-Thought Reasoning5 Oct
  2. 2025Causal Sufficiency and Necessity Improves Chain-of-Thought Reasoning11 Jun
  3. 2025Unveiling and Causalizing CoT: A Causal Pespective25 Feb
  4. 2024A Theoretical Understanding of Chain-of-Thought: Coherent Reasoning and Error-Aware Demonstration21 Oct
  5. 2023Thought Propagation: An Analogical Approach to Complex Reasoning with Large Language Models6 Oct
  6. 2023Measuring Faithfulness in Chain-of-Thought Reasoning17 Jul
  7. 2023Self-Polish: Enhance Reasoning in Large Language Models via Problem Refinement23 May
  8. 2022Large Language Models are Zero-Shot Reasoners24 May
  9. 2022Least-to-Most Prompting Enables Complex Reasoning in Large Language Models21 May
  10. 2022Self-Consistency Improves Chain of Thought Reasoning in Language Models21 Mar
  11. 2022Chain-of-Thought Prompting Elicits Reasoning in Large Language Models28 Jan

Other references · 18

  1. 2021Training Verifiers to Solve Math Word Problemscited by 9,743
  2. 2021Measuring Mathematical Problem Solving With the MATH Datasetcited by 5,203
  3. 2022Language Models (Mostly) Know What They Knowcited by 1,783
  4. 2023Large Language Models Cannot Self-Correct Reasoning Yetcited by 965
  5. 2023Question Decomposition Improves the Faithfulness of Model-Generated Reasoningcited by 107
  6. 2023Algorithm of Thoughts: Enhancing Exploration of Ideas in Large Language Modelscited by 92
  7. 2025Think-While-Generating: On-the-Fly Reasoning for Personalized Long-Form Generationcited by 29
  8. 2023Certified Robustness for Large Language Models with Self-Denoisingcited by 23
  9. 2025Measuring the Faithfulness of Thinking Drafts in Large Reasoning Modelscited by 22
  10. 2025Semantic-Space Exploration and Exploitation in RLVR for LLM Reasoningcited by 17
  11. 2025MTR-Bench: A Comprehensive Benchmark for Multi-Turn Reasoning Evaluationcited by 15
  12. 2026Double: Breaking the Acceleration Limit via Double Retrieval Speculative Parallelismcited by 15
6 more
  1. 2024Reasoning Elicitation in Language Models via Counterfactual Feedbackcited by 14
  2. 2025How Well Can Reasoning Models Identify and Recover from Unhelpful Thoughts?cited by 11
  3. 2026Spark: Strategic Policy-Aware Exploration via Dynamic Branching for Long-Horizon Agentic Learningcited by 9
  4. 2026DiffCoT: Diffusion-styled Chain-of-Thought Reasoning in LLMscited by 5
  5. 2026Unsupervised Causal Prototypical Networks for De-biased Interpretable Dermoscopy Diagnosiscited by 4
  6. 2026GCoT-Decoding: Unlocking Deep Reasoning Paths for Universal Question Answeringcited by 4

Built on by

Cited by year (2026 to 7 Oct)

Citing papers per year: 2026: 3.

3 papers.

  1. 2026Thinking With Tools, Not With Pixels: Tool Calls as Text Scaffolds for Visual Reasoningcited by 2
  2. 2026Diagnosing LLM Arbitration Behavior over Pre-evidence Epistemic States in RAG-based Fact-Checkingcited by 1
  3. 2026IDEA: An Interpretable and Editable Decision-Making Framework for LLMs via Verbal-to-Numeric Calibrationcited by 0

Similar, not cited

Close in wording, no citation either way

  1. 2026C2-Faith: Benchmarking LLM Judges for Causal and Coverage Faithfulness in Chain-of-Thought Reasoning
  2. 2025FRIT: Using Causal Importance to Improve Chain-of-Thought Faithfulness
  3. 2026Faithfulness as Information Flow: Evaluating and Training Faithful Chain-of-Thought Reasoning
  4. 2026Lie to Me: How Faithful Is Chain-of-Thought Reasoning in Reasoning Models?
  5. 2025Is Chain-of-Thought Really Not Explainability? Chain-of-Thought Can Be Faithful without Hint Verbalization
  6. 2025Unveiling Confirmation Bias in Chain-of-Thought Reasoning
  7. 2026Bypassing the Rationale: Causal Auditing of Implicit Reasoning in Language Models
  8. 2025Measuring Chain of Thought Faithfulness by Unlearning Reasoning Steps