Rui He

Conceptual disruptions in reading trigger immediate, localized processing costs, while referential disruptions unfold gradually, revealing fundamental differences in how meaning is processed by humans and LLMs.

Closed-class tokens expand and collapse in representation space faster than open-class tokens, revealing critical insights into how grammar shapes transformer geometry.

Attention might just be a cleverly disguised MLP: this work shows you can ditch the quadratic complexity and still get Transformer-level performance by dynamically predicting parameters in standard network layers.