Share your thoughts, 1 month free Claude Pro on usSee more
WorkDL logo mark

AutoPal: Autonomous Adaptation to Users for Personal AI Companionship

About

Previous research has demonstrated the potential of AI agents to act as companions that can provide constant emotional support for humans. In this paper, we emphasize the necessity of autonomous adaptation in personal AI companionship, an underexplored yet promising direction. Such adaptability is crucial as it can facilitate more tailored interactions with users and allow the agent to evolve in response to users' changing needs. However, imbuing agents with autonomous adaptability presents unique challenges, including identifying optimal adaptations to meet users' expectations and ensuring a smooth transition during the adaptation process. To address them, we devise a hierarchical framework, AutoPal, that enables controllable and authentic adjustments to the agent's persona based on user interactions. A personamatching dataset is constructed to facilitate the learning of optimal persona adaptations. Extensive experiments demonstrate the effectiveness of AutoPal and highlight the importance of autonomous adaptability in AI companionship.

Yi Cheng, Wenge Liu, Kaishuai Xu, Wenjun Hou, Yi Ouyang, Chak Tou Leong, Wenjie Li, Xian Wu, Yefeng Zheng• 2024

Related benchmarks

TaskDatasetResultRank
Distribution AlignmentPopAlign-Bench Average
Behavioral Alignment Score24
9
Distribution AlignmentPopAlign-Bench Public-A
Behav-JS25.6
9
Distribution AlignmentPopAlign-Bench Public-B
Behav-JS0.244
9
Distribution AlignmentPopAlign-Bench Reddit-B
Behav-JS0.187
9
Distribution AlignmentPopAlign-Bench Reddit-A
Behav-JS27.4
9
Dialogue Quality CalibrationPopAlign-Bench (test)
Average Score5.03
9
Showing 6 of 6 rows

Other info

Follow for update