Dianhui Chu

How Stable is the Next Token? A Geometric View of LLM Prediction Stability

PSBench: Editing Image via GUI Agents in Photoshop

VPO: Reasoning Preferences Optimization Based on $\mathcal{V}$-Usable Information

Federated Incomplete Multi-view Clustering with Globally Fused Graph Guidance

Maximizing Intermediate Checkpoint Value in LLM Pretraining with Bayesian Optimization