Long-term Task-oriented Agent: Proactive Long-term Intent Maintenance in Dynamic Environments
Qinglong Shi, Donghai Wang, Hantao Zhou, Jiguo Li, Jun Xu, Jiuchong Gao, Jinghua Hao, Renqing He
TL;DR
The paper tackles the reactive bias of current LLM-based agents by introducing proactive long-term task-oriented capabilities that maintain user intent across time in dynamic environments. It formalizes intent maintenance via Intent-Conditioned Monitoring and Event-Triggered Follow-up, and introduces ChronosBench—a 1,052-sample benchmark generated through an iterative data-synthesis pipeline—to train and evaluate these capabilities. Open-source models tuned with LoRA (e.g., Qwen3-32B) achieve up to 85.19% task completion in complex scenarios, outperforming many closed-source baselines and demonstrating the practicality of data-driven proactive dialogue for long-horizon tasks. The work provides a foundation for evaluating time-aware agent behavior in evolving environments, while acknowledging simulation-to-reality gaps and calling for further enhancements in realism and measurement of proactive assistance.
Abstract
Current large language model agents predominantly operate under a reactive paradigm, responding only to immediate user queries within short-term sessions. This limitation hinders their ability to maintain long-term user's intents and dynamically adapt to evolving external environments. In this paper, we propose a novel interaction paradigm for proactive Task-oriented Agents capable of bridging the gap between relatively static user's needs and a dynamic environment. We formalize proactivity through two key capabilities, (i) Intent-Conditioned Monitoring: The agent autonomously formulates trigger conditions based on dialog history; (ii) Event-Triggered Follow-up: The agent actively engages the user upon detecting useful environmental updates. We introduce a high-quality data synthesis pipeline to construct complex, multi-turn dialog data in a dynamic environment. Furthermore, we attempt to address the lack of evaluation criteria of task-oriented interaction in a dynamic environment by proposing a new benchmark, namely ChronosBench. We evaluated some leading close-source and open-source models at present and revealed their flaws in long-term task-oriented interaction. Furthermore, our fine-tuned model trained using synthetic data for supervised learning achieves a task completion rate of 85.19% for complex tasks including shifts in user intent, outperforming other models under test. And the result validated the effectiveness of our data-driven strategy.