mlsanityData to 7 Oct 2026

Prompting and chain-of-thought reasoning

1,002 papers7 this weekLow confidence in this grouping: the papers may belong to more than one topic, or the group may changeLow confidence in this grouping: the papers may belong to more than one topic, or the group may change · About

Keywords: chain-of-thought, distillation, reasoning, prompting, thought

Key facts. Prompting and chain-of-thought reasoning: 1,002 papers on arXiv since 2022 (data through 7 Oct 2026). 613 of its 1,002 mapped papers (61%) cite “Chain-of-Thought Prompting Elicits Reasoning in Large Language Models” (2022, arXiv 2201.11903). 339 of its 1,002 mapped papers (34%) cite “Large Language Models are Zero-Shot Reasoners” (2022, arXiv 2205.11916). 41 new papers 17 Sep–7 Oct.

What’s new

17 Sep – 7 Oct 2026

41 new papers, 17 Sep – 7 Oct; 7 this week.

  1. This week · 7
  2. Visible Reasoning Is Not a Universal Optimizer: Persona- and Thinking-Dependent Effects in Analytics Code Generation also fits: Reasoning benchmarks and evaluation
  3. Reasoning-Token Spikes Under Prompted Untruthful Responding in Large Language Models also fits: Reasoning benchmarks and evaluation
  4. Collaborative Reasoning Distillation via Cross-Feedback and Coherent Curation also fits: Reasoning benchmarks and evaluation
  5. SanSi: A Looped Typed Decision Model for System 1.5 Thinking also fits: Test-time compute and efficient thinking
  6. COMPASS: Finding Where Reasoning Lives in Language Models also fits: Reasoning benchmarks and evaluation
  7. Base Models Can Reason By Taking a Cue From Training Data also fits: Reinforcement learning for reasoning
  8. What Matters for Latent Reasoning with Flow Matching also fits: Test-time compute and efficient thinking
  9. Earlier · 34
  10. SIFT: Robust Meta-Faithfulness Verification of Chain-of-Thought Reasoning Under Distribution Shift also fits: Test-time compute and efficient thinking
  11. Learning from Repaired Reasoning: Root-Cause-Guided On-Policy Distillation also fits: Reinforcement learning for reasoning
  12. Efficient Reasoning Training Does Not Always Harm CoT Faithfulness and Monitorability
All 41
  1. On the Chain-of-Thought Monitorability of Looped Language Models
  2. Learning to Revise Reasoning with Segment-wise On-Policy Distillation also fits: Test-time compute and efficient thinking
  3. Rewarding Novel Deductions: Solver-guided Process Supervision for Logical Reasoning also fits: Test-time compute and efficient thinking
  4. When Reasoning Helps Action: Monitoring and Steering Chain-of-Thought in Vision-Language-Action Policies also fits: Reasoning for agents, planning and code
  5. When a Kindergartener Solves Calculus: Measuring Capability Leakage in Role-Prompted Reasoning Models also fits: Mathematical reasoning
  6. Diagnosing On-Policy Self-Distillation for Reasoning Language Models also fits: Test-time compute and efficient thinking
  7. Making LLMs Say What They Think: Measuring and Improving CoT-Interpretability Alignment also fits: Reasoning benchmarks and evaluation
  8. Targeted Retrieval, Compact Representations: How CoT Reasoning Improves Long-Context Counting also fits: Reasoning benchmarks and evaluation
  9. Learning from Think-Mode Advantage via On-Policy Distillation
  10. What Was Said, Not What Was 'Thought': Type-6 Logic for CoT Verification
  11. Overcoming Scaling Limits in On-Policy Self-Distillation for LLM Reasoning also fits: Reinforcement learning for reasoning
  12. Solving Without Stopping: On-Policy Distillation at Small Scale
  13. Hidden Reasoning Must Leak, but Need Not Be Readable: Fundamental Opportunities and Limits for Chain-of-Thought Monitoring
  14. Teach Yourself Where to Look: On-Policy Attention Self-Distillation for Reasoning also fits: Test-time compute and efficient thinking
  15. Understanding On-Policy Distillation: A Mechanistic Interpretability Perspective via Sparse Crosscoders also fits: Reinforcement learning for reasoning
  16. Nudgeability: Reasoning Models Follow Confidence Signals Without Tracking Their Own Competence also fits: Reasoning benchmarks and evaluation
  17. On the Token Value Inequality in Efficient Reasoning also fits: Test-time compute and efficient thinking
  18. ChemOPD: Multi-Teacher On-Policy Distillation for Multi-Task Chemical Reasoning also fits: Reasoning for agents, planning and code
  19. SeOPD: Self-Evolving LLMs via Online Policy Distillation from Self-Generated Chain-of-Thought also fits: Reasoning for agents, planning and code
  20. Learning from a Thoughtful Teacher: Adaptive On-Policy Self-Distillation for Mathematical Reasoning also fits: Mathematical reasoning
  21. Monitor Jailbreaking: Evading Chain-of-Thought Monitoring Without Encoded Reasoning
  22. Does Thinking Help Fairness? Reasoning Tokens Resolve Some Biases but Create More also fits: Causal reasoning in LLMs
  23. Reasoning Instructions Can Break Answer Decoding in Vision--Language Models
  24. Capable yet Parsimonious: Extracting and Characterizing Hidden Chain-of-Thought in Frontier Models also fits: Test-time compute and efficient thinking
  25. Spoken Language Models that Think Aloud also fits: Audio-language model reasoning
  26. From Decorative to Load-Bearing: Task Difficulty Shapes the Causal Role of Chain-of-Thought
  27. Construting Reverse Thinking: Developing Large Language Models' Reverse Thingking Ability also fits: Test-time compute and efficient thinking
  28. From Concept Alignment to Causal Grounding: An Intervention Test of Chain-of-Thought Faithfulness also fits: Causal reasoning in LLMs
  29. When Steering Fails in Latent Reasoning: A Latent-to-Language Transition Gap
  30. Learn Your Own Thoughts: Abstract Token Curriculum
  31. Chain-of-Thought Entropy as a Reliability Signal: A Preregistered Reproduction

Foundations

Most cited by papers here

  1. Chain-of-Thought Prompting Elicits Reasoning in Large Language Models2022cited by 613 here
  2. Large Language Models are Zero-Shot Reasoners2022cited by 339 here
  3. Self-Consistency Improves Chain of Thought Reasoning in Language Models2022cited by 338 here
  4. Least-to-Most Prompting Enables Complex Reasoning in Large Language Models2022cited by 179 here
  5. Language Models Don't Always Say What They Think: Unfaithful Explanations in Chain-of-Thought Prompting2023cited by 149 here

Lineage

The papers this direction builds on, oldest first.

Each paper on the line cites the one before.

  1. 2021Show Your Work: Scratchpads for Intermediate Computation with Language Modelscited by 116 here
  2. 2022Chain-of-Thought Prompting Elicits Reasoning in Large Language Modelscited by 613 here
  3. 2022Self-Consistency Improves Chain of Thought Reasoning in Language Modelscited by 338 here
  4. 2022STaR: Bootstrapping Reasoning With Reasoningcited by 104 here
  5. 2022Large Language Models are Zero-Shot Reasonerscited by 339 here
  6. 2022ReAct: Synergizing Reasoning and Acting in Language Modelscited by 83 here
  7. 2023Tree of Thoughts: Deliberate Problem Solving with Large Language Modelscited by 224 here
  8. 2024DeepSeekMath: Pushing the Limits of Mathematical Reasoning in Open Language Modelscited by 119 here
  9. 2025DeepSeek-R1: Incentivizing Reasoning Capability in LLMs via Reinforcement Learningcited by 244 here
15 more it builds on, not in this chain
  1. 2022Least-to-Most Prompting Enables Complex Reasoning in Large Language Modelscited by 179 here
  2. 2022Automatic Chain of Thought Prompting in Large Language Modelscited by 118 here
  3. 2022Challenging BIG-Bench Tasks and Whether Chain-of-Thought Can Solve Themcited by 87 here
  4. 2022Program of Thoughts Prompting: Disentangling Computation from Reasoning for Numerical Reasoning Taskscited by 87 here
  5. 2023LLaMA: Open and Efficient Foundation Language Modelscited by 88 here
  6. 2023Self-Refine: Iterative Refinement with Self-Feedbackcited by 93 here
  7. 2023Language Models Don't Always Say What They Think: Unfaithful Explanations in Chain-of-Thought Promptingcited by 149 here
  8. 2023Towards Revealing the Mystery behind Chain of Thought: A Theoretical Perspectivecited by 78 here
  9. 2023Let's Verify Step by Stepcited by 124 here
  10. 2023Measuring Faithfulness in Chain-of-Thought Reasoningcited by 141 here
  11. 2023Graph of Thoughts: Solving Elaborate Problems with Large Language Modelscited by 105 here
  12. 2024Training Large Language Models to Reason in a Continuous Latent Spacecited by 76 here
  13. 2024OpenAI o1 System Cardcited by 81 here
  14. 2025Reasoning Models Don't Always Say What They Thinkcited by 80 here
  15. 2025Qwen3 Technical Reportcited by 149 here

The 24 most cited of 2,448.

Browse

30 of 1,002, ranked 6–35

  1. 6Measuring Faithfulness in Chain-of-Thought Reasoning2023141
  2. 7Automatic Chain of Thought Prompting in Large Language Models2022118
  3. 8Graph of Thoughts: Solving Elaborate Problems with Large Language Models2023105
  4. 9Challenging BIG-Bench Tasks and Whether Chain-of-Thought Can Solve Them202287
  5. 10Program of Thoughts Prompting: Disentangling Computation from Reasoning for Numerical Reasoning Tasks202287
  6. 11Reasoning Models Don't Always Say What They Think202580
  7. 12Towards Revealing the Mystery behind Chain of Thought: A Theoretical Perspective202378
  8. 13Training Large Language Models to Reason in a Continuous Latent Space202476
  9. 14Complexity-Based Prompting for Multi-Step Reasoning202266
  10. 15Towards Understanding Chain-of-Thought Prompting: An Empirical Study of What Matters202264
All 30
  1. 16Chain of Thought Empowers Transformers to Solve Inherently Serial Problems202463
  2. 17Plan-and-Solve Prompting: Improving Zero-Shot Chain-of-Thought Reasoning by Large Language Models202363
  3. 18Large Language Models Are Reasoning Teachers202262
  4. 19Faithful Chain-of-Thought Reasoning202360
  5. 20Chain-of-Thought Reasoning In The Wild Is Not Always Faithful202556
  6. 21Teaching Small Language Models to Reason202253
  7. 22Chain of Thought Monitorability: A New and Fragile Opportunity for AI Safety202552
  8. 23The Expressive Power of Transformers with Chain of Thought202351
  9. 24Let's Think Dot by Dot: Hidden Computation in Transformer Language Models202450
  10. 25Monitoring Reasoning Models for Misbehavior and the Risks of Promoting Obfuscation202548
  11. 26Making Reasoning Matter: Measuring and Improving Faithfulness of Chain-of-Thought Reasoning202445
  12. 27Text and Patterns: For Effective Chain of Thought, It Takes Two to Tango202245
  13. 28From Explicit CoT to Implicit CoT: Learning to Internalize CoT Step by Step202444
  14. 29Navigate through Enigmatic Labyrinth A Survey of Chain of Thought Reasoning: Advances, Frontiers and Future202338
  15. 30Towards Reasoning Era: A Survey of Long Chain-of-Thought for Reasoning Large Language Models202537
  16. 31To CoT or not to CoT? Chain-of-thought helps mainly on math and symbolic reasoning202436
  17. 32CODI: Compressing Chain-of-Thought into Continuous Space via Self-Distillation202534
  18. 33Implicit Chain of Thought Reasoning via Knowledge Distillation202334
  19. 34Distilling Reasoning Capabilities into Smaller Language Models202233
  20. 35Self-Distilled Reasoner: On-Policy Self-Distillation for Large Language Models202632