ARGUS: An Agenda-Driven Framework for Reliable User Simulation

Hang Zheng, Nan Jiang, Zichen Zhu, Zixu Yang, Zhiyang Tang +4 more
10/15/2026
cs.AI

No confident code implementation found for this paper.

Abstract

High-fidelity user simulators are essential for evaluating and training agents in interactive environments. However, current LLM-based simulators struggle with reliability issues, such as goal omission and policy non-compliance. We propose ARGUS (Agenda-dRiven Goal-oriented User Simulator), a neuro-symbolic framework that maintains the user's goal state using a dynamic agenda stack manipulated via symbolic operations. By rigorously decoupling state tracking from natural language generation, ARGUS facilitates strict goal progression and constraint-aligned interactions. For cost-efficient deployment, we distill ARGUS into compact models using a unified pipeline: evaluator-guided SFT followed by online RL with dense step-level rewards. To rigorously evaluate behavioral fidelity, we introduce a rubric-based metric Composite Behavioral Score (CBS). Evaluations on $τ^3$-Bench and MultiWOZ demonstrate the significant superiority of the ARGUS architecture and the efficacy of our training pipeline. Code and data will be released soon.

View on arXivView PDF

Code Implementations

No confident code match yet

We couldn't find an author-owned or strongly-evidenced community implementation for this paper. Any repos shown below are weak matches — verify before relying on them.

No code implementations found yet.

Know of an implementation? Let us know in the comments below!

Similar Papers

Goal Alignment in LLM-Based User Simulators for Conversational AI

85% match2025

MIMESIS: Learning User Simulators as Training Environments for Interactive Agents

82% match2026

MUSE: Multi-Domain Chinese User Simulation via Self-Evolving Profiles and Rubric-Guided Alignment

82% match2026

Joint Workflow and Prompt Optimization for User Behavior Simulation

82% match2026

Goals and the Structure of Experience

81% match2025

Open-World Multi-Task Control Through Goal-Aware Representation Learning and Adaptive Horizon Prediction

81% match1 repo472026

Cite this paper

@article{zheng2026argus,
  title  = {ARGUS: An Agenda-Driven Framework for Reliable User Simulation},
  author = {Hang Zheng and Nan Jiang and Zichen Zhu and Zixu Yang and Zhiyang Tang and Shuai Fan and Lu Chen and Xiaobao Wu and Kai Yu},
  year   = {2026},
  eprint = {2610.21346},
  archivePrefix = {arXiv},
  url    = {http://arxiv.org/abs/2610.21346v1}
}

Add a badge to your README

Shows whether official code is linked for this paper.

Code of Paper badge
[![Code of Paper](https://codeofpaper.com/badge/2610.21346.svg)](https://codeofpaper.com/paper/2610.21346)

Discussion