Xiao-Tong Li

Frontier multimodal models are beginning to predict future world states, but current benchmarks reveal vast performance disparities and highlight the critical role of underlying language model capabilities.

Unlocking accurate reactive dynamics in energetic materials with MLPs trained on a mere 13,680 configurations by integrating a novel spherical-harmonic basis and knowledge-guided active learning.

A novel decoupled autoencoder design unlocks 32x spatial compression for diffusion models while simultaneously boosting reconstruction fidelity and accelerating training convergence.

RouteGraph-Mona achieves superior mineral classification accuracy by dynamically adapting routing based on sample-specific scale preferences, addressing a critical challenge in geological imaging.

AutoDesign outperforms existing design systems by aligning with human design principles and achieving superior poster generation quality through recursive self-improvement.

Executable Blender code transforms text-to-video generation, enabling unprecedented control over scene dynamics and visual fidelity.