Hua Huang

VisioMath: Benchmarking Figure-based Mathematical Reasoning in LMMs

Curvature-Guided Task Synergy for Skeleton based Temporal Action Segmentation

Discerning Minds or Generic Tutors? Evaluating Instructional Guidance Capabilities in Socratic LLMs

GeoLoom: High-quality Geometric Diagram Generation from Textual Input

LaVi: Efficient Large Vision-Language Models via Internal Feature Modulation

Revisiting Differential Attention: A Fine-Tuning Perspective on Practical Noise Mitigation

2D Quantization for Ultra‑low‑bit Optimizers

Ada-K Routing: Boosting the Efficiency of MoE-based LLMs

Rethinking Scale-Aware Temporal Encoding for Event-based Object Detection

EvFocus: Learning to Reconstruct Sharp Images from Out-of-Focus Event Streams

SubZero: Random Subspace Zeroth-Order Optimization for Memory-Efficient LLM Fine-Tuning