4 papers · 1 filter
Towards Faithful and Controllable Personalization via Critique-Post-Edit Reinforcement Learning
Chenghao Zhu, Meiling Tao, Tiannan Wang +3
Faithfully personalizing large language models (LLMs) to align with individual user preferences is a critical but challenging task. While supervised fine-tuning (SFT) quickly reach…
MiCoTA: Bridging the Learnability Gap with Intermediate CoT and Teacher Assistants
Dongyi Ding, Tiannan Wang, Chenghao Zhu +3
Large language models (LLMs) excel at reasoning tasks requiring long thought sequences for planning, reflection, and refinement. However, their substantial model size and high comp…
PersonaFeedback: A Large-scale Human-annotated Benchmark For Personalization
Meiling Tao, Chenghao Zhu, Dongyi Ding +3
With the rapid improvement in the general capabilities of LLMs, LLM personalization, i.e., how to build LLM systems that can generate personalized responses or services that are ta…
AI PERSONA: Towards Life-long Personalization of LLMs
Tiannan Wang, Meiling Tao, Ruoyu Fang +4
In this work, we introduce the task of life-long personalization of large language models. While recent mainstream efforts in the LLM community mainly focus on scaling data and com…