GRPO-RM: Fine-Tuning Representation Models via GRPO-Driven Reinforcement Learning

Fuente: arXiv
Salvato in:
Dettagli Bibliografici
Autori principali: Xu, Yanchen, Jiao, Ziheng, Zhang, Hongyuan, Li, Xuelong
Natura: Preprint
Pubblicazione: 2025
Soggetti:
Accesso online:
Tags: Aggiungi Tag
Nessun Tag, puoi essere il primo ad aggiungerne!!
_version_ 1866908665371426816
author Xu, Yanchen
Jiao, Ziheng
Zhang, Hongyuan
Li, Xuelong
author_facet Xu, Yanchen
Jiao, Ziheng
Zhang, Hongyuan
Li, Xuelong
contents The Group Relative Policy Optimization (GRPO), a reinforcement learning method used to fine-tune large language models (LLMs), has proved its effectiveness in practical applications such as DeepSeek-R1. It raises a question whether GRPO can be generalized to representation learning models. In this paper, we propose Group Relative Policy Optimization for Representation Model (GRPO-RM), and investigate the performance of GRPO-like policy in post-training representation models. Specifically, our method establishes a predefined output set to functionally replace token sequence sampling in LLMs, thereby generating an output group, which is essential for the probability-driven optimization of GRPO. In addition, a specialized reward function is designed to accommodate the properties of representation models. Extensive experiments are conducted on various real-world datasets to validate the effectiveness of our proposed method.
format Preprint
id arxiv_https___arxiv_org_abs_2511_15256
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle GRPO-RM: Fine-Tuning Representation Models via GRPO-Driven Reinforcement Learning
Xu, Yanchen
Jiao, Ziheng
Zhang, Hongyuan
Li, Xuelong
Machine Learning
Computer Vision and Pattern Recognition
The Group Relative Policy Optimization (GRPO), a reinforcement learning method used to fine-tune large language models (LLMs), has proved its effectiveness in practical applications such as DeepSeek-R1. It raises a question whether GRPO can be generalized to representation learning models. In this paper, we propose Group Relative Policy Optimization for Representation Model (GRPO-RM), and investigate the performance of GRPO-like policy in post-training representation models. Specifically, our method establishes a predefined output set to functionally replace token sequence sampling in LLMs, thereby generating an output group, which is essential for the probability-driven optimization of GRPO. In addition, a specialized reward function is designed to accommodate the properties of representation models. Extensive experiments are conducted on various real-world datasets to validate the effectiveness of our proposed method.
title GRPO-RM: Fine-Tuning Representation Models via GRPO-Driven Reinforcement Learning
topic Machine Learning
Computer Vision and Pattern Recognition
url https://arxiv.org/abs/2511.15256