HKUST

Activity Feed Request to join this org

AI & ML interests

None defined yet.

Recent Activity

ldiex submitted a paper 8 days ago

NormGuard: Reward-Preserving Norm Constraints in Flow-Matching Reinforcement Learning

Luka-Wang submitted a paper 10 days ago

LISA: Likelihood Score Alignment for Visual-condition Controllable Generation

pangly submitted a paper 25 days ago

MaskAlign: Token-Subset Representation Alignment for Efficient Diffusion Training

View all activity

Papers

NormGuard: Reward-Preserving Norm Constraints in Flow-Matching Reinforcement Learning

LISA: Likelihood Score Alignment for Visual-condition Controllable Generation

View all Papers

HKUST 's papers 13

Submitted by

Tianlin Pan

NormGuard: Reward-Preserving Norm Constraints in Flow-Matching Reinforcement Learning

2

Submitted by

Wang Yanghao

LISA: Likelihood Score Alignment for Visual-condition Controllable Generation

Submitted by

pangly

MaskAlign: Token-Subset Representation Alignment for Efficient Diffusion Training

Submitted by

Ning Lu

Policy and World Modeling Co-Training for Language Agents

3

Submitted by

JingyeChen22

Does Synthetic Layered Design Data Benefit Layered Design Decomposition?

Submitted by

Yuan

Rethinking the Diffusion Model from a Langevin Perspective

Submitted by

tianyu liu

AutoWeather4D: Autonomous Driving Video Weather Conversion via G-Buffer Dual-Pass Editing

Submitted by

Yihong Luo

TDM-R1: Reinforcing Few-Step Diffusion Models with Non-Differentiable Reward

Submitted by

Xianfeng Wu

Show, Don't Tell: Morphing Latent Reasoning into Image Generation

Submitted by

Zhitao

Dancing in Chains: Strategic Persuasion in Academic Rebuttal via Theory of Mind

Submitted by

Zipeng Wang

FlashVGGT: Efficient and Scalable Visual Geometry Transformers with Compressed Descriptor Attention

Submitted by

taesiri

GIR-Bench: Versatile Benchmark for Generating Images with Reasoning

TALE: Training-free Cross-domain Image Composition via Adaptive Latent Manipulation and Energy-guided Optimization