6 papers
Focus-then-Context: Subject-Centric Progressive Visual Token Reduction for Vision-Language Models
Yulin Zhao, Zheng Zhang
Vision-Language Models (VLMs) face a bottleneck of prohibitive computational costs arising from massive visual token sequences during inference. Existing vision token reduction met…
CachePrune: Privacy-Aware and Fine-Grained KV Cache Sharing for Efficient LLM Inference
Guanlong Wu, Zhaohan li, Yao Zhang +4
Large Language Models (LLMs) rely on Key-Value (KV) caching to accelerate inference, and many serving systems further share the KV cache across users' requests to reduce redundant…
FuRA: Full-Rank Parameter-Efficient Fine-Tuning with Spectral Preconditioning
Yequan Zhao, Ruijie Zhang, Liyan Tan +3
Both full fine-tuning (Full FT) and parameter-efficient fine-tuning methods such as LoRA introduce weight updates without accounting for the spectral structure established during p…
MemGround: Long-Term Memory Evaluation Kit for Large Language Models in Gamified Scenarios
Yihang Ding, Wanke Xia, Yiting Zhao +5
Current evaluations of long-term memory in LLMs are fundamentally static. By fixating on simple retrieval and short-context inference, they neglect the multifaceted nature of compl…
Dynamic Large Concept Models: Latent Reasoning in an Adaptive Semantic Space
Xingwei Qu, Shaowen Wang, Zihao Huang +16
Large Language Models (LLMs) apply uniform computation to all tokens, despite language exhibiting highly non-uniform information density. This token-uniform regime wastes capacity…
TreePO: Bridging the Gap of Policy Optimization and Efficacy and Inference Efficiency with Heuristic Tree-based Modeling
Yizhi Li, Qingshui Gu, Zhoufutu Wen +14
Recent advancements in aligning large language models via reinforcement learning have achieved remarkable gains in solving complex reasoning problems, but at the cost of expensive…