5 papers
Beyond SFT-to-RL: Pre-alignment via Black-Box On-Policy Distillation for Multimodal RL
Sudong Wang, Weiquan Huang, Xiaomin Yu +9
The standard post-training recipe for large multimodal models (LMMs) applies supervised fine-tuning (SFT) on curated demonstrations followed by reinforcement learning with verifiab…
BranPO: Scalable Contrastive Branch Sampling for Long-Horizon Agentic Reinforcement Learning
Yubao Zhao, Weiquan Huang, Sudong Wang +4
Agentic reinforcement learning enables large language models to perform multi-turn planning and tool use, but long-horizon training remains challenging under sparse trajectory-leve…
AMA: Adaptive Memory via Multi-Agent Collaboration
Weiquan Huang, Zixuan Wang, Hehai Lin +6
The rapid evolution of Large Language Model (LLM) agents has necessitated robust memory systems to support cohesive long-term interaction and complex reasoning. Benefiting from the…
ACE-Merging: Data-Free Model Merging with Adaptive Covariance Estimation
Bo Xu, Haotian Wu, Hehai Lin +4
Model merging aims to combine multiple task-specific expert models into a single model while preserving generalization across diverse tasks. However, interference among experts, es…
Unified-MAS: Universally Generating Domain-Specific Nodes for Empowering Automatic Multi-Agent Systems
Hehai Lin, Yu Yan, Zixuan Wang +6
Automatic Multi-Agent Systems (MAS) generation has emerged as a promising paradigm for solving complex reasoning tasks. However, existing frameworks are fundamentally bottlenecked…