De-Chuan Zhan

UniCA: Unified Covariate Adaptation for Time Series Foundation Model

Co-Evolving Latent Action World Models

Learning to Be Uncertain: Pre-training World Models with Horizon-Calibrated Uncertainty

One-Embedding-Fits-All: Efficient Zero-Shot Time Series Forecasting by a Model Zoo

Calibration Is Grouping: VR-SAG with Intra-Group Variance Control and Logit-Cluster Evaluation

Hierarchical Representation Matching for CLIP-based Class-Incremental Learning

Polaris: Scaling Up Instruction-Guided Image Generation Towards Millions of Personalized Needs

Visual Reasoning via Perceptual Extension and In-Context Learning

FrontierBench: Are We Only Testing Agents Under the Streetlight?

Dimensional Debiasing via Multi-Agent Correction

Multimodal Tabular Reasoning with Privileged Structured Information

Hawk: Leveraging Spatial Context for Faster Autoregressive Text-to-Image Generation

Leveraging Conditional Dependence for Efficient World Model Denoising

FOUNDER: Grounding Foundation Models in World Models for Open-Ended Embodied Decision Making

Parrot: Multilingual Visual Instruction Tuning

Revisiting Nearest Neighbor for Tabular Data: A Deep Tabular Baseline Two Decades Later

Rethinking Pre-Training in Tabular Data: A Neighborhood Embedding Perspective

Let the LLM Stick to Its Strengths: Learning to Route Economical LLM

Parrot: Multilingual Visual Instruction Tuning

ZooProbe: A Data Engine for Evaluating, Exploring, and Evolving Large-scale Training Data for Multimodal LLMs

Dual Consolidation for Pre-Trained Model-Based Domain-Incremental Learning

Compositional Condition Question Answering in Tabular Understanding

Addressing Imbalanced Domain-Incremental Learning through Dual-Balance Collaborative Experts

Improving LLMs for Recommendation with Out-Of-Vocabulary Tokens