Trust Your Memory: Verifiable Control of Smart Homes through Reinforcement Learning with Multi-dimensional Rewards

TL;DR

This paper introduces MemHome, a benchmark for evaluating memory-driven device control in smart homes, and releases real-world data to improve reinforcement learning methods for better memory management.

Contribution

It presents a new benchmark and dataset for memory-driven smart home control, addressing evaluation gaps and methodological challenges in reinforcement learning.

Findings

First benchmark for memory-driven device control in smart homes

Real-world dataset from anonymized user logs released

Highlights the need for fine-grained memory management in RL

Abstract

Large Language Models (LLMs) have become a key foundation for enabling personalized smart home experiences. While existing studies have explored how smart home assistants understand user queries to control devices in real time, their ability to perform memory-driven device control remains challenging from both evaluation and methodological perspectives. In terms of evaluation, existing benchmarks either focus on immediate device control or general open-domain memory retrieval tasks, and therefore cannot effectively evaluate a model's ability to perform memory-driven device control. Methodologically, while memory-driven device control can be approached using Reinforcement Learning, conventional RL methods generally rely on outcome-based supervision (i.e., whether the final task is achieved). This lack of intermediate feedback can lead to sub-optimal performance or local failures in fine-grained memory management tasks (adding, updating, deleting, and utilizing). To address these issues, we first release MemHomeLife, built from anonymized real-world long-term user interaction logs. To enable more fine-grained evaluation of different memory-related subtasks, we further construct MemHome, the first benchmark designed to systematically evaluate memory-driven device control in smart home scenarios.