Xiaoxia Wu

CARE: Covariance-Aware and Rank-Enhanced Decomposition for Enabling Multi-Head Latent Attention

Understanding and Steering the Cognitive Behaviors of Reasoning Models at Test-Time

Imitate Optimal Policy: Prevail and Induce Action Collapse in Policy Gradient

ZeRO++: Extremely Efficient Collective Communication for Large Model Training

Found in the Middle: How Language Models Use Long Contexts Better via Plug-and-Play Positional Encoding