Taesup Kim

Memory-Free Continual Learning with Null Space Adaptation for Zero-Shot Vision-Language Models

Sparse Imagination for Efficient Visual World Model Planning

Preserve and Personalize: Personalized Text-to-Image Diffusion Models without Distributional Drift

Robust Domain Generalization under Divergent Marginal and Conditional Distributions

Contrastive Residual Energy Test-time Adaptation

Patch-Level Kernel Alignment for Dense Self-Supervised Learning

PiCa: Parameter-Efficient Fine-Tuning with Column Space Projection

Towards Robust Real-World Multivariate Time Series Forecasting: A Unified Framework for Dependency, Asynchrony, and Missingness

SearchFireSafety: A Retrieval-Augmented Legal QA Dataset for Fire Safety

From Threat to Tool: Leveraging Refusal-Aware Injection Attacks for Safety Alignment

What If TSF: A Multimodal Benchmark for Conditional Time Series Forecasting with Plausible Scenarios

Language-Guided Object-Centric World Models for Predictive Control