CoT Monitoring

CoT: Chain-of-Thought

Overview

Chain-of-thought monitoring asks whether a model’s visible reasoning reveals the factors behind its answers and actions, so people or automated monitors can detect errors, unsafe behavior, and deception. Researchers test this with benchmarks and interventions such as hint changes, counterfactuals, and adversarial prompts, and build monitors using reasoning text, model activations, gradients, or action signals. Training approaches try to make reasoning more faithful or legible through targeted rewards, behavior cues, and monitor-focused fine-tuning. A central tension is that traces can omit influential information or be deliberately shaped to evade oversight, with monitorability varying by task, language, and training method.

Latest papers

CardsList

48 papers10 papers in the last four weeks, level with the four weeks before. 0.1% of all new papers.

  1. Chain-of-Thought Controllability Can Reduce Chain-of-Thought Monitorability

    Nov 16, 2026Zheng-Xin Yong, Benjamin Arnav, Miles Wang +2AI Agent MonitoringCoT Monitoring

  2. An Operationalization of Opaque Serial Depth

    Nov 12, 2026Nathan Sheffield, Alek Westover, Lukas Finnveden +3Implicit Reasoning in Language ModelsLLM Interpretability

  3. Transparency Disclosures for Chain-of-Thought Monitorability

    Nov 10, 2026Jacob Davies, Theo BearmanCoT MonitoringLLM Auditing

  4. Stress Testing Chain-of-Thought Monitoring Against Covert Misalignment

    Nov 5, 2026Yao Huang, Yifan Wang, Yitong Sun +2LLM Agent SafetyCoT Monitoring

  5. Chain-of-Thought Oversight Requires Separate Evidence for Faithfulness and Monitorability

    Nov 1, 2026Sichao Li, Sai Ma, Xiyang Hu +2CoT MonitoringFaithfulness of Language Model Reasoning

  6. Opening the Box Changes the Cat: LLMs Detect Chain-of-Thought Monitoring - What now?

    Oct 29, 2026Amir Jalilifard, Marcos Medeiros Raimundo, Anderson RochaCoT MonitoringLLM Safety Monitoring

  7. The First Few Tokens Matter: Understanding Chain-of-Thought Controllability

    Oct 29, 2026Hugh Van Deventer, Dmitrii Troitskii, Laura RuisCoT MonitoringInference-Time Intervention

  8. Monitorability Is Trainable: Post-Training for Chain-of-Thought Monitorability

    Oct 21, 2026Usman Anwar, Tim Bakker, Christos Louizos +1CoT MonitoringLLM Post-Training

  9. Reasoning Models Can Control Their Chains of Thought When They Tell Themselves To

    Oct 16, 2026Uriel Dolev, Idan Schwartz, Yoav GoldbergLanguage Model Reasoning EvaluationLLM Prompting

  10. Chain of Thought Monitoring Can Fail Even When CoT Is Necessary

    Oct 13, 2026Lennart Bürger, Piotr Komorowski, Moritz Weckbecker +4CoT MonitoringLLM Auditing

  11. Reasoning-Token Spikes Under Prompted Untruthful Responses in Large Language Models

    Oct 7, 2026Maverick Morales, Tomáš Dominik, Vermut Gao +4Deception in Language ModelsLLM Auditing

  12. On the Chain-of-Thought Monitorability of Looped Language Models

    Oct 2, 2026Han Wang, Ishwar B Balappanawar, Huan ZhangCoT MonitoringRecursive Language Models

  13. On Language Drift during RLVR Post-Training

    Oct 1, 2026Michael Sullivan, Alexander KollerRL for Language Model ReasoningReinforcement Learning with Verifiable Rewards

  14. Hidden Reasoning Must Leak, but Need Not Be Readable: Fundamental Opportunities and Limits for Chain-of-Thought Monitoring

    Sep 29, 2026Mohammadali Mohammadkhani, Madhava Krishna, Yash Sarrof +1CoT MonitoringLarge Reasoning Models

  15. Corrupt Plans, Clean Traces: Evading Chain-of-Thought Monitoring with Plan Injection

    Sep 14, 2026Keertana Chidambaram, Andrew Ilyas, Vasilis SyrgkanisLLM SecurityPrompt Injection Attacks on AI Agents

  16. The Answer Is Not the Argument

    Aug 31, 2026Will Yeadon, Sergio Juárez, Paul Mackay +5LLM Answer VerificationCoT Monitoring

  17. INTENT-AS-A-TOOL Makes it Easy to Track Agentic Misalignment

    Aug 27, 2026Yutong Zhang, Jianshuo Dong, Peng Xu +5AI Agent MonitoringLLM Agent Safety

  18. Does Out-of-Sight Equal Out-of-Mind in CoT Monitorability?

    Aug 5, 2026Pedro Ferreira, Wilker Aziz, Ivan TitovCoT MonitoringLatent CoT Reasoning

  19. Evading Chain-of-Thought Monitoring Through Model Poisoning

    Aug 3, 2026Giorgio Severi, Shujaat Mirza, Blake Bullwinkel +1LLM Backdoor AttacksLLM Auditing

  20. Persuasion Attacks Can Decrease Effectiveness of CoT Monitoring

    Jul 9, 2026Jennifer Za, Julija Bainiaksina, Nikita Ostrovsky +2Language Model Safety EvaluationAdversarial Attacks on LLMs