Jierun Chen

From Verifiable Dot to Reward Chain: Harnessing Verifiable Reference-based Rewards for Reinforcement Learning of Open-ended Generation

InSight-o3: Empowering Multimodal Foundation Models with Generalized Visual Search

ModalMix: Optimizing Multimodal Data Mixtures with Compute-Dependent Regression

DualRPO: All-in-one Visual RL with Internal and External Rewards

Revisiting Referring Expression Comprehension Evaluation in the Era of Large Multimodal Models