11 papers
Route to Remember: Adaptive Memory Selection for LLM Agents
Zeyu Fang, Jun Yin, Yongsheng Mei +8
The memory of production agents is increasingly becoming a bottleneck and is built almost exclusively from text: key points are extracted from out-of-limit context and retrieved in…
MELD: Recursive Self-Improvement via Hierarchical Coordination for Multi-Agent Scientific Discovery
Sizhe Tang, Zuyuan Zhang, Zeyu Fang +1
Multi-agent scientific discovery with large language models raises a coupled learning problem: research assignments determine both which experiments are pursued and the experiences…
Differentiable Policy Transport over Multi-Layer Network Feasibility Geometry
Zuyuan Zhang, Zeyu Fang, Mahdi Imani +2
Learning-based control is increasingly central to automating network operations. A learned policy, however, must satisfy cross-layer constraints on interference, power-rate couplin…
IntentScore: Intent-Conditioned Action Evaluation for Computer-Use Agents
Rongqian Chen, Yu Li, Zeyu Fang +3
Computer-Use Agents (CUAs) leverage large language models to execute GUI operations on desktop environments, yet they generate actions without evaluating action quality, leading to…
ZODIAC: Zero-shot Offline Diffusion for Inferring Multi-xApps Conflicts in Open Radio Access Networks
Zeyu Fang, Shu Hong, Huu Trung Thieu +2
Open Radio Access Network (O-RAN) enables network control through multi-vendor xApps operating both within and across layers, subnets, and domains, whose concurrent execution can t…
Reasoning Knowledge-Gap in Drone Planning via LLM-based Active Elicitation
Zeyu Fang, Beomyeol Yu, Cheng Liu +5
Human-AI joint planning in Unmanned Aerial Vehicles (UAVs) typically relies on control handover when facing environmental uncertainties, which is often inefficient and cognitively…
Knowing When to Ask: Resolving Uncertainty in Human-Robot Joint Planning via Explicit Dialogue and Implicit Intent Cues
Zeyu Fang, Yuxin Lin, Cheng Liu +6
Effective human-robot collaboration in open-world environments requires joint planning under uncertainty about the task, the environment, and the human teammate. Communication is t…
Structuring Value Representations via Geometric Coherence in Markov Decision Processes
Zuyuan Zhang, Zeyu Fang, Tian Lan
Geometric properties can be leveraged to stabilize and speed reinforcement learning. Existing examples include encoding symmetry structure, geometry-aware data augmentation, and en…
Manifold-Constrained Energy-Based Transition Models for Offline Reinforcement Learning
Zeyu Fang, Zuyuan Zhang, Mahdi Imani +1
Model-based offline reinforcement learning is brittle under distribution shift: policy improvement drives rollouts into state--action regions weakly supported by the dataset, where…
MINT: Minimal Information Neuro-Symbolic Tree for Objective-Driven Knowledge-Gap Reasoning and Active Elicitation
Zeyu Fang, Mahdi Imani, Tian Lan
Joint planning through language-based interactions is a key area of human-AI teaming. Planning problems in the open world often involve various aspects of incomplete information an…
Learning from Random Demonstrations: Offline Reinforcement Learning with Importance-Sampled Diffusion Models
Zeyu Fang, Tian Lan
Generative models such as diffusion have been employed as world models in offline reinforcement learning to generate synthetic data for more effective learning. Existing work eithe…