8 papers
DLLG: Dynamic Logit-Level Gating of LLM Experts
Bingnan Li, Zhaoyang Zhang, Xiaoze Liu +6
Leveraging multiple specialized LLMs can combine complementary strengths, but existing approaches trade adaptability for stability: routing commits prematurely, heuristic ensemblin…
Do Proactive Agents Really Need an LLM to Decide When to Wake and What to Anchor?
Xiaoze Liu, Ruowang Zhang, Amir H. Abdi +5
Proactive agents read user activity as text and call an LLM on every event to decide whether to act. But user activity is not natively text: it is a structured event stream of (act…
The Vision Wormhole: Latent-Space Communication in Heterogeneous Multi-Agent Systems
Xiaoze Liu, Ruowang Zhang, Weichen Yu +7
Multi-Agent Systems (MAS) powered by Large Language Models have unlocked advanced collaborative reasoning, yet they remain bottlenecked by discrete text communication, which impose…
When the Same Coefficients Reach Different Places: Asymmetric Realizability in Transplanting Tokenizers across Large Language Models
Xiaoze Liu, Weichen Yu, Matt Fredrikson +2
Tokenizer transplant in cross-vocabulary model composition reconstructs donor-only embedding rows as weighted combinations over shared lexical anchors and reuses those coefficients…
SUV: Scalable Large Language Model Copyright Compliance with Regularized Selective Unlearning
Tianyang Xu, Xiaoze Liu, Feijie Wu +2
Large Language Models (LLMs) have transformed natural language processing by learning from massive datasets, yet this rapid progress has also drawn legal scrutiny, as the ability t…
Towards Federated RLHF with Aggregated Client Preference for LLMs
Feijie Wu, Xiaoze Liu, Haoyu Wang +3
Reinforcement learning with human feedback (RLHF) fine-tunes a pretrained large language model (LLM) using user preference data, enabling it to generate content aligned with human…