13 papers
LLMVisor: A Real-Time Latency Attribution Model for Multi-Tenant LLM Serving
Shuowei Jin, Xueshen Liu, Jiaxin Shan +4
As LLM inference shifts to multi-tenant GPU clusters, co-batching improves throughput but obscures per-tenant usage and limits control. Enabling fractional sharing of the inference…
HeaPA: Difficulty-Aware Heap Sampling and On-Policy Query Augmentation for LLM Reinforcement Learning
Weiqi Wang, Xin Liu, Binxuan Huang +13
RLVR has become a standard recipe for training LLMs on reasoning tasks with verifiable outcomes, but when rollout generation dominates the cost, efficiency hinges on which prompts…
CoMem: Context Management with A Decoupled Long-Context Model
Yuwei Zhang, Chengyu Dong, Shuowei Jin +11
Context management enables agentic models to solve long-horizon tasks through iterative summarization of previous interaction histories. However, this process typically incurs subs…
AstraFlow: Dataflow-Oriented Reinforcement Learning for Agentic LLMs
Haizhong Zheng, Yizhuo Di, Jiahui Wang +7
Reinforcement learning (RL) is increasingly used to improve the reasoning, coding, and tool-use capabilities of large language models, but agentic RL remains prohibitively expensiv…
VIP: Visual-guided Prompt Evolution for Efficient Dense Vision-Language Inference
Hao Zhu, Shuo Jin, Wenbin Liao +4
Pursuing training-free open-vocabulary semantic segmentation in an efficient and generalizable manner remains challenging due to the deep-seated spatial bias in CLIP. To overcome t…
Learning with Rare Success but Rich Feedback via Reflection-Enhanced Self-Distillation
Yuwei Zhang, Sha Li, Changlong Yu +9
Enabling Large Language Models (LLMs) to continuously improve from environmental interactions is a central challenge in post-training. While on-policy self-distillation offers a pr…