Cohere Labs at COLM: Chain-of-Thought Legibility Is Not Real Interpretability
Cohere_Labs · x · 2026-10-06
- Cohere Labs presented "Legibility is Not Interpretability: Comparing Judged and Actual Importance in Chain-Of-Thought Reasoning" at COLM.
- Key finding: how legible or salient a CoT step appears does not match its actual causal importance to the answer — written reasoning can diverge from the model's real basis.
- The work was done by intern Kevin Du; Julia Kreutzer also contributed posters and a workshop talk.
More from Research
- Huawei Noah's Tail-Influence Sampling Cuts CVaR Policy Evaluation MSE by Up to 76% — huawei-noah · 2026-10-06
- Google's KeyRec Achieves Best Long-Video VLM Results With Just 10% of Visual Token Budget — google · 2026-10-06
- 4DCodeBench Shows Frontier Models Reconstruct Static Scenes but Fail at Dynamics — 4DCodeBench · 2026-10-06
- OmniTaskonomy: Year-long study shows generation training can improve understanding tasks — WeijiaShi2 · 2026-10-06
- Newton proved the product rule without limits, using a discrete symmetric-difference trick — ctjlewis · 2026-10-06
- DeepMind's AI designs enzymes from scratch: 99x drug building block yield, plastic-eating at 90°C — 141_1337 · 2026-10-06