Yu Lu
2 papers ยท Latest:
Computer Vision
Power Reinforcement Post-Training of Text-to-Image Models with Super-Linear Advantage Shaping
SLAS improves text-to-image models by using a novel super-linear advantage shaping to mitigate reward hacking and enhance training efficiency and robustness.
2605.10937
Machine LearningParameter Importance is Not Static: Evolving Parameter Isolation for Supervised Fine-Tuning
EPI dynamically isolates critical parameters during SFT, reducing interference and forgetting by adapting to evolving parameter importance.
2604.14010
๐ฌ Weekly AI Paper Digest
Get the top 10 AI/ML arXiv papers from the week โ summarized, scored, and delivered to your inbox every Monday.