What a chain of thought is evidence of

A decodable answer, a difficulty threshold, an unmarked important step, and an argument about the word "reasoning".

1. Kyle Cox, Darius Kianersi, and Adria Garriga-Alonso, "Post-Hoc Reasoning in Chain of Thought: Decoding and Steering Pre-Committed Answers," arXiv:2603.01437 (2026). https://arxiv.org/abs/2603.01437
2. Siddharth Boppana, Annabel Ma, Max Loeffler, Raphael Sarfati, Eric Bigelow, Atticus Geiger, Owen Lewis, and Jack Merullo, "Reasoning Theater: Disentangling Model Beliefs from Chain-of-Thought," arXiv:2603.05488 (2026). https://arxiv.org/abs/2603.05488
3. Kevin Du, Alexander Hoyle, Laura Ruis, and Acyr Locatelli, "Legibility is Not Interpretability: Comparing Judged and Actual Importance in Chain-Of-Thought Reasoning," arXiv:2609.04194 (2026). https://arxiv.org/abs/2609.04194
4. Subbarao Kambhampati, Kaya Stechly, Karthik Valmeekam, Lucas Saldyt, Siddhant Bhambri, and colleagues, "Position: Stop Anthropomorphizing Intermediate Tokens as Reasoning/Thinking Traces!", arXiv:2504.09762 (2025). https://arxiv.org/abs/2504.09762