SSFO: Self-Supervised Faithfulness Optimization for Retrieval-Augmented Generation

Fuente: arXiv
Guardado en:
Detalles Bibliográficos
Autores principales: Tang, Xiaqiang, Wang, Yi, Hu, Keyu, Xu, Rui, Li, Chuang, Sun, Weigao, Li, Jian, Xie, Sihong
Formato: Preprint
Publicado: 2025
Materias:
Acceso en línea:
Etiquetas: Agregar Etiqueta
Sin Etiquetas, Sea el primero en etiquetar este registro!
_version_ 1866912628948860928
author Tang, Xiaqiang
Wang, Yi
Hu, Keyu
Xu, Rui
Li, Chuang
Sun, Weigao
Li, Jian
Xie, Sihong
author_facet Tang, Xiaqiang
Wang, Yi
Hu, Keyu
Xu, Rui
Li, Chuang
Sun, Weigao
Li, Jian
Xie, Sihong
contents Retrieval-Augmented Generation (RAG) systems require Large Language Models (LLMs) to generate responses that are faithful to the retrieved context. However, faithfulness hallucination remains a critical challenge, as existing methods often require costly supervision and post-training or significant inference burdens. To overcome these limitations, we introduce Self-Supervised Faithfulness Optimization (SSFO), the first self-supervised alignment approach for enhancing RAG faithfulness. SSFO constructs preference data pairs by contrasting the model's outputs generated with and without the context. Leveraging Direct Preference Optimization (DPO), SSFO aligns model faithfulness without incurring labeling costs or additional inference burden. We theoretically and empirically demonstrate that SSFO leverages a benign form of \emph{likelihood displacement}, transferring probability mass from parametric-based tokens to context-aligned tokens. Based on this insight, we propose a modified DPO loss function to encourage likelihood displacement. Comprehensive evaluations show that SSFO significantly outperforms existing methods, achieving state-of-the-art faithfulness on multiple context-based question-answering datasets. Notably, SSFO exhibits strong generalization, improving cross-lingual faithfulness and preserving general instruction-following capabilities. We release our code and model at the anonymous link: https://github.com/chkwy/SSFO
format Preprint
id arxiv_https___arxiv_org_abs_2508_17225
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle SSFO: Self-Supervised Faithfulness Optimization for Retrieval-Augmented Generation
Tang, Xiaqiang
Wang, Yi
Hu, Keyu
Xu, Rui
Li, Chuang
Sun, Weigao
Li, Jian
Xie, Sihong
Computation and Language
Artificial Intelligence
Retrieval-Augmented Generation (RAG) systems require Large Language Models (LLMs) to generate responses that are faithful to the retrieved context. However, faithfulness hallucination remains a critical challenge, as existing methods often require costly supervision and post-training or significant inference burdens. To overcome these limitations, we introduce Self-Supervised Faithfulness Optimization (SSFO), the first self-supervised alignment approach for enhancing RAG faithfulness. SSFO constructs preference data pairs by contrasting the model's outputs generated with and without the context. Leveraging Direct Preference Optimization (DPO), SSFO aligns model faithfulness without incurring labeling costs or additional inference burden. We theoretically and empirically demonstrate that SSFO leverages a benign form of \emph{likelihood displacement}, transferring probability mass from parametric-based tokens to context-aligned tokens. Based on this insight, we propose a modified DPO loss function to encourage likelihood displacement. Comprehensive evaluations show that SSFO significantly outperforms existing methods, achieving state-of-the-art faithfulness on multiple context-based question-answering datasets. Notably, SSFO exhibits strong generalization, improving cross-lingual faithfulness and preserving general instruction-following capabilities. We release our code and model at the anonymous link: https://github.com/chkwy/SSFO
title SSFO: Self-Supervised Faithfulness Optimization for Retrieval-Augmented Generation
topic Computation and Language
Artificial Intelligence
url https://arxiv.org/abs/2508.17225