PsyBer-Agent: A Psychology-Driven User Agent for Recommender Systems
Summary
The paper proposes PsyBer-Agent, a psychology-inspired LLM user simulator for recommender systems that models evolving latent states like fatigue and shifting preferences, calibrated via Unbalanced Gromov-Wasserstein optimal transport, along with the WebSim platform and PsyBer Benchmark for evaluating behavioral fidelity at scale.
View Cached Full Text
Cached at: 10/14/26, 06:50 AM
Paper page - PsyBer-Agent: A Psychology-Driven User Agent for Recommender Systems
Source: https://huggingface.co/papers/2610.13989
Abstract
Evaluatingrecommendersystemsbeforedeploymentrequiressimulatorsthatremainrealisticoverlong,closed-loopinteractioncycles.CurrentLLM-basedagentsgenerallyfunctionassurfaceimitatorsthatgeneraterealisticshort-termresponsesbutfailtocapturetheunderlyinguserdynamicssuchasfatigue,affect,andshiftingpreferences.WeproposePsyBer-Agent,apsychology-inspiredsimulatorthatmodelsthesedynamicsthroughitsPsy-Engine.Thisenginerepresentseachuserwiththreeevolving,model-definedlatentstates.Tocalibratesimulatedbehavioragainstrealinteractionlogs,wealignlog-conditionedreferencetrajectorieswithsimulator-generatedlatent-statetrajectoriesviaUnbalancedGromov--Wassersteinoptimaltransport.Thisapproachprovidesstructuredsupervisionforalightweightpolicywithoutthecomputationalburdenoffine-tuningthebackboneLLM.WealsointroduceWebSimandthePsyBerBenchmark,whichencompassmultimodalscenariosandsevenrecommenderbackbones.Ourframeworkevaluatesbehavioralfidelity,latent-stateconsistency,robustness,andresponsestosocial-cueperturbations.Onthetestedhorizons,PsyBer-Agentshowslowerbehavioraldiscrepancythantheprompt-onlysimulatorswhilemaintainingstableperformanceunderlarge-scaleconcurrentsimulationforupto5Mqueuedsimulationtasksunderboundedconcurrency.Codeisavailableinthesupplementarymaterialandat:PsyBer-Agent:https://anonymous.4open.science/r/CC66F/,WebSimPlatform:https://anonymous.4open.science/r/D8EAX/.
View arXiv pageView PDFGitHub2Add to collection
Get this paper in your agent:
hf papers read 2610\.13989
Don’t have the latest CLI?curl \-LsSf https://hf\.co/cli/install\.sh \| bash
Models citing this paper0
No model linking this paper
Cite arxiv.org/abs/2610.13989 in a model README.md to link it from this page.
Datasets citing this paper0
No dataset linking this paper
Cite arxiv.org/abs/2610.13989 in a dataset README.md to link it from this page.
Spaces citing this paper0
No Space linking this paper
Cite arxiv.org/abs/2610.13989 in a Space README.md to link it from this page.
Collections including this paper0
No Collection including this paper
Add this paper to acollectionto link it from this page.
Similar Articles
Prompt Optimization for User Simulation in Conversational Recommender Systems: A Multi-Objective Framework
This paper proposes a framework to automatically optimize prompts for LLM-based user simulators in conversational recommender systems, addressing issues like positive bias and limited behavioral diversity.
Auto-RecSys: Harnessing Autonomous Research Agents for Industry-Scale Recommender System
Auto-RecSys is an autonomous research system for automating long-horizon experimentation on industry-scale recommendation models, using distributed execution, centralized memory, and cognitive-procedural separation to improve efficiency and reliability.
Beyond Cooperative Simulators: Generating Realistic User Personas for Robust Evaluation of LLM Agents
Introduces Persona Policies (PPol), a plug-and-play control layer that uses LLM-driven evolutionary program search to generate diverse, human-like user personas for evaluating LLM agents. Achieves 33–62% fitness gains over baseline, with human-likeness rated at 80.4%, and improves agent robustness with +17% task success.
auto-psych: Automating the science of mind using agent-driven theory discovery and experimentation
auto-psych is an agent-based system that automates theory discovery and experimentation in computational cognitive science, using LLM agents to generate hypotheses, design experiments, and analyze data from crowdsourced participants. It demonstrates faster and better theory generation compared to human-derived theories in a classic psychology paradigm.
Poor Man's Agentic Modeling: Simulating Large LLM-Agent Societies on a Laptop
This paper proposes a method for simulating large LLM-agent societies on a laptop by fitting low-parameter surrogate models from a few hundred queries, using a statistical-physics-based taxonomy to predict when this approximation holds. The approach is validated on EconAgent and several other simulations using DeepSeek-elicited agent behaviors.