Goal Alignment in LLM-Based User Simulators for Conversational AI

TL;DR

This paper introduces UGST, a framework for improving goal alignment in LLM-based user simulators, enhancing their reliability in multi-turn conversational AI applications.

Contribution

We propose UGST, a novel framework for tracking user goals in LLM-based simulators, along with a three-stage development methodology and new evaluation metrics.

Findings

Significant improvements in goal alignment on MultiWOZ 2.4 and τ-Bench benchmarks.

Enhanced ability of simulators to track and reason about user goals.

Addressed a key gap in the reliability of goal-oriented user simulation.

Abstract

User simulators are essential to conversational AI, enabling scalable agent development and evaluation through simulated interactions. While current Large Language Models (LLMs) have advanced user simulation capabilities, we reveal that they struggle to consistently demonstrate goal-oriented behavior across multi-turn conversations--a critical limitation that compromises their reliability in downstream applications. We introduce User Goal State Tracking (UGST), a novel framework that tracks user goal progression throughout conversations. Leveraging UGST, we present a three-stage methodology for developing user simulators that can autonomously track goal progression and reason to generate goal-aligned responses. Moreover, we establish comprehensive evaluation metrics for measuring goal alignment in user simulators, and demonstrate that our approach yields substantial improvements across two benchmarks (MultiWOZ 2.4 and {\tau}-Bench). Our contributions address a critical gap in conversational AI and establish UGST as an essential framework for developing goal-aligned user simulators.