Cheng Jin

Large Language Model Compression with Global Rank and Sparsity Optimization

Pref-GRPO: Pairwise Preference Reward-based GRPO for Stable Text-to-Image Reinforcement Learning

Unified Reward Model for Multimodal Understanding and Generation

Complete Structure Guided Point Cloud Completion via Cluster- and Instance-Level Contrastive Learning

Towards Robust Influence Functions with Flat Validation Minima

Computational Budget Should Be Considered in Data Selection

Unified Multimodal Chain-of-Thought Reward Model through Reinforcement Fine-Tuning

Out-of-Distribution Detection using Neural Activation Prior