Saved in:
Bibliographic Details
Main Authors: Chen, Zekai, Daniel, Weeden, Chen, Po-yu, Buet-Golfouse, Francois
Format: Preprint
Published: 2024
Subjects:
Online Access:https://arxiv.org/abs/2404.16115
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866913328696131584
author Chen, Zekai
Daniel, Weeden
Chen, Po-yu
Buet-Golfouse, Francois
author_facet Chen, Zekai
Daniel, Weeden
Chen, Po-yu
Buet-Golfouse, Francois
contents The advent of personalized content generation by LLMs presents a novel challenge: how to efficiently adapt text to meet individual preferences without the unsustainable demand of creating a unique model for each user. This study introduces an innovative online method that employs neural bandit algorithms to dynamically optimize soft instruction embeddings based on user feedback, enhancing the personalization of open-ended text generation by white-box LLMs. Through rigorous experimentation on various tasks, we demonstrate significant performance improvements over baseline strategies. NeuralTS, in particular, leads to substantial enhancements in personalized news headline generation, achieving up to a 62.9% improvement in terms of best ROUGE scores and up to 2.76% increase in LLM-agent evaluation against the baseline.
format Preprint
id arxiv_https___arxiv_org_abs_2404_16115
institution arXiv
publishDate 2024
record_format arxiv
spellingShingle Online Personalizing White-box LLMs Generation with Neural Bandits
Chen, Zekai
Daniel, Weeden
Chen, Po-yu
Buet-Golfouse, Francois
Computation and Language
Artificial Intelligence
Machine Learning
The advent of personalized content generation by LLMs presents a novel challenge: how to efficiently adapt text to meet individual preferences without the unsustainable demand of creating a unique model for each user. This study introduces an innovative online method that employs neural bandit algorithms to dynamically optimize soft instruction embeddings based on user feedback, enhancing the personalization of open-ended text generation by white-box LLMs. Through rigorous experimentation on various tasks, we demonstrate significant performance improvements over baseline strategies. NeuralTS, in particular, leads to substantial enhancements in personalized news headline generation, achieving up to a 62.9% improvement in terms of best ROUGE scores and up to 2.76% increase in LLM-agent evaluation against the baseline.
title Online Personalizing White-box LLMs Generation with Neural Bandits
topic Computation and Language
Artificial Intelligence
Machine Learning
url https://arxiv.org/abs/2404.16115