4 papers
PReCache: Efficient KV Cache Sharing for Multi-LoRA Agents via Low-Rank Precomputation and Neutral Reconstruction
Hyesung Jeon, Hyeongju Ha, Jae-Joon Kim
Multi-LoRA agent systems enable efficient role specialization by sharing a common backbone model. However, each agent repeatedly processes the growing shared trajectory and constru…
WaveFront Decoding: Parallelized Self-Speculative Decoding for Looped Language Models
Hyeongju Ha, Jae-Joon Kim
Looped language models repeatedly apply a weight-shared block to increase effective depth without increasing parameter count, but the resulting T sequential recurrent-block calls p…
RelayGen: Intra-Generation Model Switching for Efficient Reasoning
Jiwon Song, Yoongon Kim, Jae-Joon Kim
Large reasoning models (LRMs) achieve strong performance on complex reasoning tasks by generating long, multi-step reasoning trajectories, but inference-time scaling incurs substan…
LRAgent: Efficient KV Cache Sharing for Multi-LoRA LLM Agents
Hyesung Jeon, Hyeongju Ha, Jae-Joon Kim
Role specialization in multi-LLM agent systems is often realized via multi-LoRA, where agents share a pretrained backbone and differ only by lightweight adapters. Despite sharing b…