Xiao Yan

SynthCoder: Anti-pattern identification and model training for FIM mode code completion

This work introduces SynthCoder, which adopts optimized masking strategies better aligned with developer’s expectation in FIM, and first refine AST-level node masking and add heuristics that better mimic developers’ expectation to construct the training corpora.

SynthCoder: Anti-pattern identification and model training for FIM mode code completion

BEE: Intervention-Adaptive Real-World Reinforcement Learning with Vision-Language-Action Models

BEE is introduced, an intervention-adaptive framework for real-world RL on a frozen VLA that lets the policy go BEyond Expert imitation, and a Correction Model predicts how a human would correct a given VLA proposal and how consistent the correction is along each action dimension.

CIDER: Continual Interactive Distillation for Embodied Reinforcement Learning

This work introduces Continual Interactive Distillation for Embodied Reinforcement Learning (CIDER), a continual reinforcement learning framework that freezes the accumulated historical policy as a teacher before learning each new task and interleaves task learning with distillation-based retention.

VINE: Taming Generative Control Policies for Reinforcement Learning

VINE is proposed, an RL-oriented sampling method that enables stable end-to-end value-gradient optimization for flow-matching policies and achieves stable policy improvement and consistently outperforms state-of-the-art RL methods on the OGBench offline RL benchmark and real-world robotic manipulation task.