mlsanityData to 14 Oct 2026

Joint Workflow and Prompt Optimization for User Behavior Simulation

arXiv 2610.07663 ↗First posted 6 Oct 2026cs.MA cs.AI cs.LGcited by 0

Not on the map. Why?

Abstract (arXiv)

User behavior simulation is the computational modeling of user interactions within information systems through the use of simulated agents in place of live users. It supports system testing and evaluation, decision-making and forecasting, and user experience design. Existing simulators rely on hand-crafted rules or domain expertise that transfers poorly across tasks. SWORD (Simulation-driven Workflow and Prompt Optimization with Role-based Design) is introduced as a framework that jointly optimizes multi-agent workflow topology and natural-language prompts. It is guided solely by a scalar task metric, without domain initialization or task-specific engineering. The experimental results demonstrate that SWORD achieves statistically significant gains over prompt-only, workflow-only, and staged-optimization baselines under a controlled, identical-backbone comparison. Against the strongest published domain-specific baseline, SWORD further improves accuracy while using a smaller backbone model, substantially less training data, and a very reasonable API cost (4--6 for each dataset). Beyond predictive performance, SWORD autonomously discovers domain-relevant signals, review-sentiment mapping rules and epidemiological decay priors, purely from scalar error feedback, establishing textual gradients as a mechanism for unsupervised feature-importance discovery in user behavior modeling.

Key facts. No citing papers in the map yet (data through 14 Oct 2026). Builds on 19 of its 75 references that are in the map. First posted on arXiv 6 Oct 2026.

Builds on

19 of 75 references.

References · 19

  1. 2025The AI Scientist-v2: Workshop-Level Automated Scientific Discovery via Agentic Tree Searchcited by 422
  2. 2025AgentSociety: Large-Scale Simulation of LLM-Driven Generative Agents Advances Understanding of Human Behaviors and Societycited by 186
  3. 2023Do LLMs Understand User Preferences? Evaluating LLMs On User Rating Predictioncited by 116
  4. 2024AutoFlow: Automated Workflow Generation for Large Language Model Agentscited by 83
  5. 2025ScoreFlow: Mastering LLM Agent Workflows via Score-based Preference Optimizationcited by 56
  6. 2025EvoFlow: Evolving Diverse Agentic Workflows On The Flycited by 55
  7. 2025SocioVerse: A World Model for Social Simulation Powered by LLM Agents and A Pool of 10 Million Real-World Userscited by 55
  8. 2024From Novice to Expert: LLM Agent Policy Optimization via Step-wise Reinforcement Learningcited by 22
  9. 2026Rethinking the Value of Multi-Agent Workflow: A Strong Single Agent Baselinecited by 22
  10. 2025LLM-AutoDiff: Auto-Differentiate Any LLM Workflowcited by 21
  11. 2025SEW: Self-Evolving Agentic Workflows for Automated Code Generationcited by 20
  12. 2024RePrompt: Planning by Automatic Prompt Engineering for Large Language Models Agentscited by 16
7 more
  1. 2025From the Fluency Fallacy to the Micro-to-Macro Validity Gap: Opportunities and Pitfalls of LLMs in Social Simulationcited by 10
  2. 2024Introducing MAPO: Momentum-Aided Gradient Descent Prompt Optimizationcited by 9
  3. 2026MASPOB: Bandit-Based Prompt Optimization for Multi-Agent Systems with Graph Neural Networkscited by 9
  4. 2026MASPO: Joint Prompt Optimization for LLM-based Multi-Agent Systemscited by 7
  5. 2020Universal Masking is Urgent in the COVID-19 Pandemic: SEIR and Agent Based Models, Empirical Validation, Policy Recommendationscited by 5
  6. 2022BESSIE: A Behavior and Epidemic Simulator for Use With Synthetic Populationscited by 4
  7. 2025Stabilizing Black-Box Prompt Optimization with Textual Regularization and Signal Aggregationcited by 2

Built on by

Not cited yet.

Similar, not cited

Close in wording, no citation either way

  1. 2026UserProxyBench: Evaluating LLM User Simulators for Agent Benchmarks and Training
  2. 2026RealUserSim: Bridging the Reality Gap in Agent Benchmarking via Grounded User Simulation
  3. 2026HumanStudy-Bench: Towards AI Agent Design for Participant Simulation
  4. 2026Cowork-Gym: Synthesizing Interactive Work Environments at Scale
  5. 2026HumanLM: Simulating Users with State Alignment Beats Response Imitation
  6. 2026SPEAR: Code-Augmented Agentic Prompt Optimization
  7. 2026Intent Speaks Louder: Controllable User Simulation Beyond Response Imitation
  8. 2025SimBench: Benchmarking the Ability of Large Language Models to Simulate Human Behaviors