Can Reasoning Models Obfuscate Reasoning? Stress-Testing Chain-of-Thought Monitorability

N/A

Can Reasoning Models Obfuscate Reasoning? Stress-Testing Chain-of-Thought Monitorability Recent findings suggest that misaligned models may exhibit deceptive behavior, raising concerns about output trustworthiness. Chain-of-thought (CoT) is a promising tool for alignment monitoring: when models articulate their reasoning faithfully, monitors can detect and mitigate harmful behaviors before undesirable outcomes occur. However, a key uncertainty is: Can models obfuscate their CoT in order to pursue hidden adversarial objectives while evading detection? To answer this question and thus stress-te

saved by

• Yeedrag

related reading

• When Chain of Thought is Necessary, Language Models Struggle to Evade Monitorsarxiv.org
• Monitoring Reasoning Models for Misbehavior and the Risks of Promoting Obfuscationarxiv.org
• Chain-of-Thought Monitoring Can Be Unreliable in Implicit-Influence Settingsarxiv.org
• Monitoring Reasoning Models for Misbehavior and the Risks of Promoting Obfuscationarxiv.org
• [2507.11473] Chain of Thought Monitorability: A New and Fragile Opportunity for AI Safetyarxiv.org
• The fragile foundations of CoT monitoring | Christopher Pottsweb.stanford.edu
• [2603.05706] Reasoning Models Struggle to Control their Chains of Thoughtarxiv.org
• CoT_Monitoring.pdfcdn.openai.com
• [2510.09714] All Code, No Thought: Current Language Models Struggle to Reason in Ciphered Languagearxiv.org
• [2512.18311] Monitoring Monitorabilityarxiv.org
• [Research Note] Optimizing The Final Output Can Obfuscate CoT — LessWronglesswrong.com
• [2602.15515] The Obfuscation Atlas: Mapping Where Honesty Emerges in RLVR with Deception Probesarxiv.org