LLM Agent Workflow Optimization

LLM: Large Language Model

This field studies how to organize language models, tools, memory, and decision steps so agents can complete multi-step tasks effectively and efficiently. Approaches include fixed workflows and DAGs, query-level model routing, reinforcement learning and evolutionary search, and trace-guided prompt or harness refinement; they are tested in settings such as coding, web use, retrieval, and scientific research. Systems work also optimizes execution through caching, scheduling, parallelism, and compiling repeated behavior into deterministic workflows. A central tension is whether adaptive workflows improve on reusable fixed designs enough to justify their added cost and complexity, while maintaining reliability and generalizing beyond the tasks used for optimization.

Latest papers

262 papers59 papers in the last four weeks, up 69% on the four weeks before. 0.3% of all new papers.

Locating and Closing Information Gaps in Multi-Agent Systems
Oct 29, 2026Yaxin Gong, Chongming Gao, Gangyi Zhang +7LLM Agent Workflow Optimization Multi-Agent Systems
University of Science and Technology of China · Qwen Business Unit of Alibaba · National University of Singapore
Learning to Formulate: Agentic Hardware-Software Co-Design of AI Infrastructure through Progressive Optimization and Verification
Oct 29, 2026Yi Ding, Li ShangHardware-Software Co-Design AI Accelerator Inference
Shenzhen Loop Area Institute (SLAI) · National University of Singapore
Evolving Agent Teams
Oct 27, 2026Shiyi Cao, Ziming Mao, Dacheng Li +2Multi-Agent Coding Self-Evolving Agents
University of California, Berkeley
A Harness Reducing over-Defensive Writing in Agents by Isolating Objections
Oct 27, 2026Xiaochen ChangLLM Agent Harnesses LLM Agent Workflow Optimization
RouteDRAG: Learning to Route and Control Agentic Execution for Multimodal Document RAG
Oct 27, 2026Zekai Chen, Hao Zhu, Yu Liu +5Agentic RAG Multimodal RAG
Beijing Institute of Technology, Beijing, China · Chinese Academy of Sciences, China · Zhejiang University, Hangzhou, China +3
CohortFlow: A Steerable Clinical Cohort Curation Agent with Localized Semantic Correction
Oct 26, 2026Neil Chen, Adrian Serapio, Hanxue Gu +3LLM Agent Workflow Optimization LLM Agent Efficiency
UC Berkeley · UC San Francisco
Work Agents: Systems, Capabilities, and Optimization
Oct 26, 2026Zishan Xu, Enjun Du, Mingqi Gao +13LLM Agent Reliability LLM Agent Workflow Optimization
Shanghai Jiao Tong University · The University of Hong Kong · The Hong Kong University of Science and Technology (Guangzhou) +5
AIPC: Evidence-Guided Agent Workflows for PyTorch Project Deployment
Oct 26, 2026Zhanwei Wu, Jianhao Su, Brian Zhang +11AI Accelerator Inference LLM Agent Workflow Optimization
Qualcomm Technologies, Inc · Independent Researcher
CodeMeepo: A Multi-Agent Workflow for Repository-Level Code Migration and Optimization
Oct 25, 2026Farzad Habibi, Changrong Wu, Jinzhou Jiang +1Code Translation Multi-Agent Coding
University of California, Irvine · Microsoft · OpenAI
EMMA: Structured Episodic Memory for Multi-Agent Workflow Synthesis
Oct 25, 2026Ruijia Niu, Kaiwen Liu, Rose Yu +1Multi-Agent LLM Systems LLM Agent Memory
University of California San Diego · Amazon
MatOS: An Agent Operating System for Materials Research
Oct 25, 2026Dingyang Lyu, Jing Gao, Zhonghao Chang +13Autonomous Research Agents LLM Agent Workflow Optimization
Zhongguancun Academy · Zhongguancun Institute of Artificial Intelligence, Beijing, China
FlashScript: Compiling Recurring Agent Workflows into Reusable Neuro-Symbolic Programs
Oct 24, 2026Sadra Sabouri, Hosein Hasanbeig, Adam Sage +6LLM Agent Workflow Optimization LLM-Based Program Synthesis
University of Southern California, Los Angeles, CA · Microsoft, Redmond, WA
AgentOrch: How Close Can Global Orchestration Come to Step-by-Step Interaction?
Oct 24, 2026Zijie Liu, Qi Yang Huang, Jinhao Duan +6LLM Agent Orchestration Enterprise AI Agents
University of North Carolina at Chapel Hill · Workato
Adaptive Skill Granularity for Coding Agents: Runtime Routing and Macro-to-Micro Recovery
Oct 22, 2026Dingxuan Zhang, Yanhui Jiang, Jun Wang +1Coding Agents LLM Agent Workflow Optimization
Shanghai Jiao Tong University, China · Kuaishou Technology Co., Ltd., China · University of Warwick, UK
AgentDP: Towards Automated Model Deployment for Heterogeneous Embedded AI Applications
Oct 22, 2026Leming Shen, Qiang Yang, Xinyu Huang +2MLOps On-Device Inference
The Hong Kong Polytechnic University · University College London · U2University of Cambridge
Mining the Design Space of Multi-Agent Workflows
Oct 22, 2026Xavier Cadet, Junyan Cheng, Peter Chin +1Multi-Agent Orchestration LLM Agent Workflow Optimization
Dartmouth College · Allen Institute for AI
RoleFlow: Self-Evolving Multi-Role Agentic Collaboration via Structural Adaptation
Oct 21, 2026Yi Ge, Pan Wang, Xuehao Cui +4Multi-Agent LLM Systems LLM Agent Workflow Optimization
University of Maryland, College Park · University of Pittsburgh · The University of Texas at Dallas
PAUSE: Search-Free Steering of Coding Agents Before Their First Source-Code Edit
Oct 21, 2026Viral Patel, Boyin Tan, Jason Xue +1Coding Agents LLM Agent Workflow Optimization
Mohamed bin Zayed University of Artificial Intelligence, Abu Dhabi, United Arab Emirates
Recursive Self-Improvement on a Budget: Scoring Code Edits Before They Run
Oct 21, 2026Erfan Darzi, Sree Bhargavi Balija, Sheng HeLLM Agent Self-Improvement LLM Agent Workflow Optimization
Harvard University · Massachusetts Institute of Technology · UC San Diego
Which Model for Which Role? Cost-Efficient Pareto Recommendation for LLM Agentic Workflows
Oct 19, 2026Qian Xie, Wenyue Hua, Yueli He +7LLM Agent Workflow Optimization Multi-Objective Optimization
Columbia University · Cornell University · Microsoft +2
Behavioral Probes for Information Flow in LLM Swarms
Oct 19, 2026Junhui Chen, Yu Zhang, Lantian Li +1Multi-Agent LLM Systems LLM Agent Workflow Optimization
School of Artificial Intelligence, Beijing University of Posts and Telecommunications, China · Center for Speech and Language Technologies, BNRist, Tsinghua University, China
IF-MAP: Influence-Function-Inspired Agent Prioritization in Multi-Agent Systems
Oct 17, 2026Jie Mao, Pingchen Lu, Zikun Qu +6Multi-Agent LLM Systems LLM Agent Workflow Optimization
The Chinese University of Hong Kong, Shenzhen · Harbin Institute of Technology · School of Engineering and Applied Sciences, Harvard University
Who Plans, Who Edits? Verified, Cost-Aware Division of Labor for LLM-Driven Analog Layout Repair
Oct 16, 2026Xilai Dai, Saeed MohammadiLLM Agents Electronic Design Automation
Elmore Family School of Electrical and Computer Engineering, Purdue University, West Lafayette, IN 47907, USA
TokenEconomizer: Experience-Driven Black-Box Cost Control for Coding Agents
Oct 15, 2026Aaron J. Li, Weijia Xu, Chandan Singh +8Token Efficiency Coding Agents
UC Berkeley · Microsoft Research · Microsoft Security
FAOPs: Normal Forms and an Algebra for AI Agent Workflows
Oct 15, 2026Ali Mohammadi ShanghooshabadLLM Agent Workflow Optimization Runtime Enforcement for AI Agents
JPMorgan Chase & Co.
RSIFlow: Adaptive Self-Improvement Workflows with an Evolving Meta-Agent
Oct 15, 2026Huan Zhang, Mingju Chen, Dongxu Zhou +4LLM Agent Self-Improvement Self-Evolving Agents
Beijing Advanced Innovation Center for Future Blockchain and Privacy Computing, School of Artificial Intelligence, Beihang University · Tsinghua University · Beijing Academy of Artificial Intelligence (BAAI)
Jev-LiteMAS: Separating Decisions from Generation in Multi-Agent Workflows
Oct 15, 2026Kaixiang Wang, Zhaijiacheng Zhou, Jiong Lou +1Multi-Agent LLM Systems LLM Agent Orchestration
Shanghai Jiao Tong University, Shanghai, China
JevCoder: When Does Decision Delegation Pay Off in Code Repair?
Oct 14, 2026Jiahao Zhang, Yifan Zhang, Yu HuangAutomated Program Repair Coding Agents
Vanderbilt University, Nashville, Tennessee, USA
When Is Smaller Greener? Right-Sizing Language Models for Agentic AI
Oct 14, 2026Sajid Siraj, Maria BarbatiLLM Agent Workflow Optimization LLM Inference Efficiency
Leeds University Business School, University of Leeds, Leeds, United Kingdom · Ca’Foscari University of Venice, Venice, Italy
The Last Mile of Coding Agents for Long-Document Tasks: An Empirical Study on Clinical Trial Protocol Digitization
Oct 13, 2026Jiacheng Lin, Geoffrey Low, Nishit Agarwal +4Coding Agents Document Information Extraction
University of Illinois Urbana-Champaign · Medidata Solutions

Oct 29, 2026J/K move · Enter open · S save

Locating and Closing Information Gaps in Multi-Agent Systems

Yaxin Gong, Chongming Gao, Gangyi Zhang, Chenxiao Fan+6

University of Science and Technology of China · Qwen Business Unit of Alibaba · National University of Singapore

LLM Agent Workflow Optimization Multi-Agent Systems Multi-Agent Communication Multi-Agent Communication Topology LLM Agent Reliability

Open paper AskPDF arXiv

Multi-step Web tasks often require several LLM agents to gather and combine evidence. The communication graph of a fixed workflow restricts which agents can read each other's outputs: evidence retrieved by one agent may never reach the agent that needs it, leaving an information gap and an incomplete or incorrect answer. Adding a connection can deliver the missing evidence, but can also lead the receiver to adopt an incorrect upstream conclusion that displaces a correct answer; the final output alone does not separate the two. We propose Selective Probing and Link Insertion for Closing information gaps in multi-agent Execution (SPLICE), which locates information gaps per query and decides what to adopt after adding an edge. After the original workflow runs, SPLICE uses the query and node outputs to locate a possible gap, adds one directed edge from a source to a receiver to deliver the relevant evidence, and re-executes the receiver and affected downstream nodes, producing a candidate answer. The adoption decision reads both the final answers of the two runs and the process changes the edit caused inside the workflow: how far the receiver moved, where its new content came from, and whether the change reached the final answer. A selector trained on these process features chooses between adopting the candidate answer and retaining the original. On five benchmarks, SPLICE raises the task score on Web tool use by 8.0% and the mean score on four reasoning and planning benchmarks by 5.7% relative to the static skeleton, at 1.25 to 1.44

\times

the skeleton's cost, and transfers zero-shot to four unseen backbones.