Xiangxin Zhu — accepted AI-conference papers

2026

iGRPO: Fast Online RL for Flow Matching Model with Dense Reward

Sucheng Ren, Chen Chen, Zhenbang Wang, Liangchen Song, Xiangxin Zhu, Yinfei Yang +1

ICML 2026poster

Conventional practice assumes that online reinforcement learning for flow-matching models requires sampling full denoising trajectories to compute rewards. This assumption underlies methods such as Group Relative Policy Optimization (GRPO), where the policy must traverse the entire reverse process b…

Cited by 0 Source ↗Scholar ↗

2025

CLIP-UP: A Simple and Efficient Mixture-of-Experts CLIP Training Recipe with Sparse Upcycling

Xinze Wang, Chen Chen, Yinfei Yang, Hong-You Chen, Bowen Zhang, Aditya Pal +2

EMNLP 2025

Mixture-of-Experts (MoE) models are crucial for scaling model capacity while controlling inference costs. While integrating MoE into multimodal models like CLIP improves performance, training these models is notoriously challenging and expensive. We propose CLIP-Upcycling (CLIP-UP), an efficient alt

Cited by 0 Source ↗PDF ↗Scholar ↗