Han Hu

5 papers Beijing Institute of Technology

• Efficient Federated Learning against Byzantine Attacks and Data Heterogeneity via Aggregating Normalized Gradients
• Merging on the Fly Without Retraining: A Sequential Approach to Scalable Continual Model Merging
• MixPrompt: Efficient Mixed Prompting for Multimodal Semantic Segmentation
• RBench-V: A Primary Assessment for Visual Reasoning Models with Multimodal Outputs
• Self-Evolving Pseudo-Rehearsal for Catastrophic Forgetting with Task Similarity in LLMs