E3-Agent: An Executable and Evolving Agent for Resource Management of Edge Generative Inference
On the map: LLM agents and tool use → General LLM agent frameworks and benchmarks
Abstract (arXiv)
Edge deployments of generative inference increasingly face two practical realities: per-device per-model performance is often unknown at deployment time, and it is non-stationary due to user-driven semantic events, background load, and device churn. Consequently, a resource manager that is tuned offline under a fixed regime can become brittle and expensive to maintain. This paper presents E3-Agent, an executable and evolving agent for edge artificial intelligence generated content (AIGC) resource management. E3-Agent separates a fast-path router that makes millisecond-level dispatch decisions from a slow-path, event-driven large language model (LLM) meta-controller that mitigates regime shifts through a small, explicit control surface exposed via a tool interface, including risk gating, router configuration, and rapid performance calibration. The agent learns online from execution feedback and continuously adapts to unknown and time-varying service-time mappings. We evaluate E3-Agent in a discrete-event simulator driven by MLPerf-derived device-model measurement priors, covering cold-start warmup and three dynamic regimes: semantic dynamics, device churn, and hidden drift. Across the dynamic scenarios, E3-Agent reduces average latency by 65%-73% compared to the best static baseline, stays within 7%-10% of an online full-information Oracle used for evaluation, and effectively suppresses stutter rate under semantic degradation.
Key facts. No citing papers in the map yet (data through 7 Oct 2026). Placed in General LLM agent frameworks and benchmarks, one of 16 directions in LLM agents and tool use. Builds on 14 of its 24 references that are in the map. First posted on arXiv 21 May 2026.
Builds on
14 of 24 references.
References · 14
- 2022ReAct: Synergizing Reasoning and Acting in Language Modelscited by 8,355
- 2023Efficient Memory Management for Large Language Model Serving with PagedAttentioncited by 5,767
- 2023SDXL: Improving Latent Diffusion Models for High-Resolution Image Synthesiscited by 4,290
- 2024SWE-agent: Agent-Computer Interfaces Enable Automated Software Engineeringcited by 1,628
- 2024DistServe: Disaggregating Prefill and Decoding for Goodput-optimized Large Language Model Servingcited by 595
- 2019MLPerf Inference Benchmarkcited by 195
- 2024Parrot: Efficient Serving of LLM-based Applications with Semantic Variablecited by 96
- 2024BurstGPT: A Real-world Workload Dataset to Optimize LLM Serving Systemscited by 84
- 2024Overview of AI and Communication for 6G Network: Fundamentals, Challenges, and Future Research Opportunitiescited by 25
- 2024Hybrid SD: Edge-Cloud Collaborative Inference for Stable Diffusion Modelscited by 11
- 2024Towards Wireless Native Big AI Model: The Mission and Approach Differ From Large Language Modelcited by 7
- 2025Edge Agentic AI Framework for Autonomous Network Optimisation in O-RANcited by 4
Built on by
Not cited yet.
Similar, not cited
Close in wording, no citation either way
- 2026Multi-Agentic AI for Fairness-Aware and Accelerated Multi-modal Large Model Inference in Real-world Mobile Edge Networks
- 2026From PyTorch to the NPU: LLM-Agent-Driven Model Conversion Across Heterogeneous Inference Runtimes
- 2026A Self-Calibrating Agentic AI Framework for Autonomous Edge Resource Allocation
- 2026End-to-End Latency-Minimizing and Load-Balanced Request Scheduling for Edge LLM Inference in Agentic AI Services
- 2026EdgeFM: Efficient Edge Inference for Vision-Language Models
- 2025Cognitive Edge Computing: A Comprehensive Survey on Optimizing Large Models and AI Agents for Pervasive Deployment
- 2026Evaluating Inference Compute for Generative AI: A Framework for Enterprise Workloads
- 2026Agentic Performance at the Edge: Insights from Benchmarking