How do LLMs do CoT reasoning internally? In our new #ACL2026 paper, we show that reasoning unfolds as a structured trajectory in representation space. Correct and incorrect paths diverge, and we use this to predict correctness before the answer and correct errors mid-flight. 1/

💡New paper! Woke up to Anthropic's emotion paper and realized “wait, that's our finding too.” We concurrently uncovered a circular valence & arousal (VA) geometry of emotions, steering refusal & sycophancy. We further provide a mechanistic account: tokens occupy distinct regions in this space. 1/

🚨New #ACL2025 paper! Today’s “safe” language models can look unbiased—but alignment can actually make them more biased implicitly by reducing their sensitivity to race-related associations. 🧵Find out more below!