Pengkun Wang

StreamMTS: Towards Streaming Multivariate Time Series Forecasting

Current mainstream research in multivariate time series (MTS) forecasting often assumes that all data are static. However, real-world MTS data typically arrives continuously in a streaming manner, which we refer to as streaming MTS. The statistical characteristics and spatiotemporal graph topology of these data evolve...

LAMP: Long-Tailed Multi-modal Prompt Tuning for Vision-Language Models

DiDPO: Diff-in-Diff Policy Optimization for Coding Agent Training

DiDPO is proposed, a critic-free RL method that constructs fine-grained credit units directly from the structure of code diffs directly from the structure of code diffs, offering a principled framework for fine-grained credit assignment in coding agent training.