4 papers
RedKnot-MLA: Multi-Head Offline-Online Reuse for DeepSeek-V4 Long-Context Serving
Yang Liu, Zhaokai Luo, Huayi Jin +10
Multi-head latent attention (MLA) exposes many logical query heads through one packed latent KV stream. This representation is memory efficient, but it removes the physical per-hea…
RedKnot: Efficient Long-Context LLM Serving with Head-Aware KV Reuse and SegPagedAttention
Yang Liu, Zhaokai Luo, Huayi Jin +6
As the input length of large language model (LLM) serving continues to grow, the KV cache has become a dominant bottleneck in AI infrastructure. It limits GPU memory capacity, serv…
EvoAgent: An Evolvable Agent Framework with Skill Learning and Multi-Agent Delegation
Aimin Zhang, Jiajing Guo, Fuwei Jia +3
This paper proposes EvoAgent--an evolvable large language model (LLM) agent framework that integrates structured skill learning with a hierarchical sub-agent delegation mechanism.…
LVP-CLIP:Revisiting CLIP for Continual Learning with Label Vector Pool
Yue Ma, Huantao Ren, Boyu Wang +3
Continual learning aims to update a model so that it can sequentially learn new tasks without forgetting previously acquired knowledge. Recent continual learning approaches often l…