# Lost in Simulation: LLM-Simulated Users are Unreliable Proxies for Human Users in Agentic Evaluations

arXiv: https://arxiv.org/abs/2601.17087 · HTML: https://mlsanity.com/paper/2601.17087/

## Key facts

- Cited by 40 papers in the map (data through 14 Oct 2026).
- Placed in Social simulation and multi-agent debate, one of 16 directions in LLM agents and tool use.
- Builds on 34 of its 47 references that are in the map.
- First posted on arXiv 23 Jan 2026.

## Placed in

- LLM agents and tool use / [Social simulation and multi-agent debate](https://mlsanity.com/agents/social-simulation-and-multi-agent-debate/)

## Builds on (most cited first, 30 of 34)

- [On Calibration of Modern Neural Networks](https://mlsanity.com/paper/1706.04599/) (2017)
- [WebArena: A Realistic Web Environment for Building Autonomous Agents](https://mlsanity.com/paper/2307.13854/) (2023)
- [τ-bench: A Benchmark for Tool-Agent-User Interaction in Real-World Domains](https://mlsanity.com/paper/2406.12045/) (2024)
- [gpt-oss-120b & gpt-oss-20b Model Card](https://mlsanity.com/paper/2508.10925/) (2025)
- [Whose Opinions Do Language Models Reflect?](https://mlsanity.com/paper/2303.17548/) (2023)
- [Towards Measuring the Representation of Subjective Global Opinions in Language Models](https://mlsanity.com/paper/2306.16388/) (2023)
- [TravelPlanner: A Benchmark for Real-World Planning with Language Agents](https://mlsanity.com/paper/2402.01622/) (2024)
- [TheAgentCompany: Benchmarking LLM Agents on Consequential Real World Tasks](https://mlsanity.com/paper/2412.14161/) (2024)
- [Annotators with Attitudes: How Annotator Beliefs And Identities Bias Toxic Language Detection](https://mlsanity.com/paper/2111.07997/) (2021)
- [Two Tales of Persona in LLMs: A Survey of Role-Playing and Personalization](https://mlsanity.com/paper/2406.01171/) (2024)
- [MultiChallenge: A Realistic Multi-Turn Conversation Evaluation Benchmark Challenging to Frontier LLMs](https://mlsanity.com/paper/2501.17399/) (2025)
- [Survey of Cultural Awareness in Language Models: Text and Beyond](https://mlsanity.com/paper/2411.00860/) (2024)
- [NLPositionality: Characterizing Design Biases of Datasets and Models](https://mlsanity.com/paper/2306.01943/) (2023)
- [Collaborative Gym: A Framework for Enabling and Evaluating Human-Agent Collaboration](https://mlsanity.com/paper/2412.15701/) (2024)
- [AI Suggestions Homogenize Writing Toward Western Styles and Diminish Cultural Nuances](https://mlsanity.com/paper/2409.11360/) (2024)
- [Measurement to Meaning: A Validity-Centered Framework for AI Evaluation](https://mlsanity.com/paper/2505.10573/) (2025)
- [Investigating African-American Vernacular English in Transformer-Based Text Generation](https://mlsanity.com/paper/2010.02510/) (2020)
- [SimulatorArena: Are User Simulators Reliable Proxies for Multi-Turn Evaluation of AI Assistants?](https://mlsanity.com/paper/2510.05444/) (2025)
- [Evaluating Large Language Models as Generative User Simulators for Conversational Recommendation](https://mlsanity.com/paper/2403.09738/) (2024)
- [Exploring Cross-Cultural Differences in English Hate Speech Annotations: From Dataset Construction to Analysis](https://mlsanity.com/paper/2308.16705/) (2023)
- [Evaluating Cultural and Social Awareness of LLM Web Agents](https://mlsanity.com/paper/2410.23252/) (2024)
- [Sociodemographic Prompting is Not Yet an Effective Approach for Simulating Subjective Judgments with LLMs](https://mlsanity.com/paper/2311.09730/) (2023)
- [ChatBench: From Static Benchmarks to Human-AI Evaluation](https://mlsanity.com/paper/2504.07114/) (2025)
- [Can LLM Agents Simulate Multi-Turn Human Behavior? Evidence from Real Online Customer Behavior Data](https://mlsanity.com/paper/2503.20749/) (2025)
- [ECom-Bench: Can LLM Agent Resolve Real-World E-commerce Customer Support Issues?](https://mlsanity.com/paper/2507.05639/) (2025)
- [Survey of Social Bias in Vision-Language Models](https://mlsanity.com/paper/2309.14381/) (2023)
- [Goal Alignment in LLM-Based User Simulators for Conversational AI](https://mlsanity.com/paper/2507.20152/) (2025)
- [High-Dimension Human Value Representation in Large Language Models](https://mlsanity.com/paper/2404.07900/) (2024)
- [GRASP: A Disagreement Analysis Framework to Assess Group Associations in Perspectives](https://mlsanity.com/paper/2311.05074/) (2023)
- [Real or Robotic? Assessing Whether LLMs Accurately Simulate Qualities of Human Responses in Dialogue](https://mlsanity.com/paper/2409.08330/) (2024)

## Built on by (most cited first, 30 of 40)

- [Mind the Sim2Real Gap in User Simulation for Agentic Tasks](https://mlsanity.com/paper/2603.11245/) (2026)
- [RealUserSim: Bridging the Reality Gap in Agent Benchmarking via Grounded User Simulation](https://mlsanity.com/paper/2605.20204/) (2026)
- [SWE-INTERACT: Reimagining SWE Benchmarks as User-Driven Long-Horizon Coding Sessions](https://mlsanity.com/paper/2606.30573/) (2026)
- [Beyond Cooperative Simulators: Generating Realistic User Personas for Robust Evaluation of LLM Agents](https://mlsanity.com/paper/2605.12894/) (2026)
- [FutureSim: Replaying World Events to Evaluate Adaptive Agents](https://mlsanity.com/paper/2605.15188/) (2026)
- [ThoughtTrace: Understanding User Thoughts in Real-World LLM Interactions](https://mlsanity.com/paper/2605.20087/) (2026)
- [Measuring and Mitigating the Distributional Gap Between Real and Simulated User Behaviors](https://mlsanity.com/paper/2605.07847/) (2026)
- [Bridging the Agent-World Gap: Text World Models for LLM-based Agents](https://mlsanity.com/paper/2606.09032/) (2026)
- [Reinforcing Human Behavior Simulation via Verbal Feedback](https://mlsanity.com/paper/2605.20506/) (2026)
- [AppWorld-UL: Benchmarking Diverse Agent-User Interactions for Tool-Use](https://mlsanity.com/paper/2607.20536/) (2026)
- [Multi-Agent Home Energy Management Assistant](https://mlsanity.com/paper/2602.15219/) (2026)
- [Cognitive Models and AI Algorithms Provide Templates for Designing Language Agents](https://mlsanity.com/paper/2602.22523/) (2026)
- [Learning to Interrupt in Language-based Multi-agent Communication](https://mlsanity.com/paper/2604.06452/) (2026)
- [Auditing Support Strategies in LLMs through Grounded Multi-Turn Social Simulation](https://mlsanity.com/paper/2604.17079/) (2026)
- [VISTA: A Versatile Interactive User Simulation Toolkit for Agent Evaluation](https://mlsanity.com/paper/2606.11079/) (2026)
- [Simulated Customers Never Walk Away: Decision Fidelity of LLM User Simulators Measured Against Real Purchase Outcomes](https://mlsanity.com/paper/2606.20708/) (2026)
- [Can LLMs Think Like Consumers? Benchmarking Crowd-Level Reaction Reconstruction with ConsumerSimBench](https://mlsanity.com/paper/2605.17079/) (2026)
- [CRAB-Bench: Evaluating LLM Agents under Complex Task Dependencies and Human-aligned User Simulation](https://mlsanity.com/paper/2606.01815/) (2026)
- [Distribution-First Population Simulation: Collapse, Calibration, and Recall in Non-WEIRD LLM Persona Modeling](https://mlsanity.com/paper/2607.18310/) (2026)
- [Measurement Without Validity: The Compounding Reliability Problem in Agentic AI Evaluation](https://mlsanity.com/paper/2608.00794/) (2026)
- [AcCoRD: Evaluating User-Agent Collaboration Under Realistic User Preference Dynamics](https://mlsanity.com/paper/2608.27818/) (2026)
- [Disclosure-Gated User Simulation for Companion-Agent Evaluation](https://mlsanity.com/paper/2609.00982/) (2026)
- [Agent-Integrated Software: Interaction Contracts and Continuous Assurance](https://mlsanity.com/paper/2609.11381/) (2026)
- [UserProxyBench: Evaluating LLM User Simulators for Agent Benchmarks and Training](https://mlsanity.com/paper/2609.38043/) (2026)
- [The textitSilicon Society Cookbook: Design Space of LLM-based Social Simulations](https://mlsanity.com/paper/2605.00197/) (2026)
- [SEATauBench: Progressively Adapting Tool-Agent-User Evaluation Into Low-Resource Southeast Asian Languages](https://mlsanity.com/paper/2606.28715/) (2026)
- [AppLooper: An Agentic Application Engineering Loop for Accountable Release with Virtual-User Feedback](https://mlsanity.com/paper/2608.14093/) (2026)
- ["Act Like a 5th Grader" is Not Enough: Bounding Knowledge in LLM-Based User Simulators](https://mlsanity.com/paper/2608.30033/) (2026)
- [Investigating Assistant Bias in LLM User Simulators Using a Role Vector](https://mlsanity.com/paper/2609.00608/) (2026)
- [Evaluating Open-Weight E-Commerce Agents with Environment-Grounded Verification](https://mlsanity.com/paper/2609.16093/) (2026)

## Abstract (arXiv)

Agentic benchmarks increasingly rely on LLM-simulated users to scalably evaluate agent performance, yet the robustness, validity, and fairness of this approach remain unexamined. Through a user study with participants across the United States, India, Kenya, and Nigeria, we investigate whether LLM-simulated users serve as reliable proxies for real human users in evaluating agents on τ-Bench retail tasks. We find that user simulation lacks robustness, with agent success rates varying up to 9 percentage points across different user LLMs. Furthermore, evaluations using simulated users exhibit systematic miscalibration, underestimating agent performance on challenging tasks and overestimating it on moderately difficult ones. African American Vernacular English (AAVE) speakers experience consistently worse success rates and calibration errors than Standard American English (SAE) speakers, with disparities compounding significantly with age. We also find simulated users to be a differentially effective proxy for different populations, performing worst for AAVE and Indian English speakers. Additionally, simulated users introduce conversational artifacts and surface different failure patterns than human users. These findings demonstrate that current evaluation practices risk misrepresenting agent capabilities across diverse user populations and may obscure real-world deployment challenges.