Fine-Grained Verifiers: Preference Modeling as Next-token Prediction in Vision-Language Alignment

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Cui, Chenhang, Zhang, An, Zhou, Yiyang, Chen, Zhaorun, Deng, Gelei, Yao, Huaxiu, Chua, Tat-Seng
Format: Preprint
Published: 2024
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866916697311543296
author Cui, Chenhang
Zhang, An
Zhou, Yiyang
Chen, Zhaorun
Deng, Gelei
Yao, Huaxiu
Chua, Tat-Seng
author_facet Cui, Chenhang
Zhang, An
Zhou, Yiyang
Chen, Zhaorun
Deng, Gelei
Yao, Huaxiu
Chua, Tat-Seng
contents The recent advancements in large language models (LLMs) and pre-trained vision models have accelerated the development of vision-language large models (VLLMs), enhancing the interaction between visual and linguistic modalities. Despite their notable success across various domains, VLLMs face challenges in modality alignment, which can lead to issues like hallucinations and unsafe content generation. Current alignment techniques often rely on coarse feedback and external datasets, limiting scalability and performance. In this paper, we propose FiSAO (Fine-Grained Self-Alignment Optimization), a novel self-alignment method that utilizes the model's own visual encoder as a fine-grained verifier to improve vision-language alignment without the need for additional data. By leveraging token-level feedback from the vision encoder, FiSAO significantly improves vision-language alignment, even surpassing traditional preference tuning methods that require additional data. Through both theoretical analysis and experimental validation, we demonstrate that FiSAO effectively addresses the misalignment problem in VLLMs, marking the first instance of token-level rewards being applied to such models.
format Preprint
id arxiv_https___arxiv_org_abs_2410_14148
institution arXiv
publishDate 2024
record_format arxiv
spellingShingle Fine-Grained Verifiers: Preference Modeling as Next-token Prediction in Vision-Language Alignment
Cui, Chenhang
Zhang, An
Zhou, Yiyang
Chen, Zhaorun
Deng, Gelei
Yao, Huaxiu
Chua, Tat-Seng
Computer Vision and Pattern Recognition
Computation and Language
The recent advancements in large language models (LLMs) and pre-trained vision models have accelerated the development of vision-language large models (VLLMs), enhancing the interaction between visual and linguistic modalities. Despite their notable success across various domains, VLLMs face challenges in modality alignment, which can lead to issues like hallucinations and unsafe content generation. Current alignment techniques often rely on coarse feedback and external datasets, limiting scalability and performance. In this paper, we propose FiSAO (Fine-Grained Self-Alignment Optimization), a novel self-alignment method that utilizes the model's own visual encoder as a fine-grained verifier to improve vision-language alignment without the need for additional data. By leveraging token-level feedback from the vision encoder, FiSAO significantly improves vision-language alignment, even surpassing traditional preference tuning methods that require additional data. Through both theoretical analysis and experimental validation, we demonstrate that FiSAO effectively addresses the misalignment problem in VLLMs, marking the first instance of token-level rewards being applied to such models.
title Fine-Grained Verifiers: Preference Modeling as Next-token Prediction in Vision-Language Alignment
topic Computer Vision and Pattern Recognition
Computation and Language
url https://arxiv.org/abs/2410.14148