Score-Control for Hallucination Reduction in Diffusion Models

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Bhosale, Mahesh, Devulapally, Naresh Kumar, Wasi, Abdul, Pham, Chau, Lokhande, Vishnu Suresh, Doermann, David
Format: Preprint
Published: 2026
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866913174954967040
author Bhosale, Mahesh
Devulapally, Naresh Kumar
Wasi, Abdul
Pham, Chau
Lokhande, Vishnu Suresh
Doermann, David
author_facet Bhosale, Mahesh
Devulapally, Naresh Kumar
Wasi, Abdul
Pham, Chau
Lokhande, Vishnu Suresh
Doermann, David
contents Diffusion models have emerged as the backbone of modern generative AI, powering advances in vision, language, audio and other modalities. Despite their success, they suffer from hallucinations, implausible samples that lie outside the support of true data distribution, which degrade reliability and trust. In this work, we first empirically confirm previously proposed hypothesis that score smoothness causes hallucinations in Image Generation diffusion models and provide a density-based perspective. We further formalize this notion by linking the hallucinations probability mass to lipschitz constant of the learned score function. Motivated by this, we introduce a Variance-Guided Score Modulation (VSM) strategy that controls the score Jacobian, in turn reducing score smoothness and better approximating the ground truth score that decreases hallucinations. Empirical results on synthetic and real-world datasets demonstrate that our approach reduces hallucinations (up to ~25%) while maintaining high fidelity and diversity, providing a principled step toward more reliable diffusion-based image generation. We also propose two benchmark datasets with extreme semantic variation for systematic hallucination evaluation. Code and Datasets are publicly available at https://github.com/bhosalems/VSM.
format Preprint
id arxiv_https___arxiv_org_abs_2606_00377
institution arXiv
publishDate 2026
record_format arxiv
spellingShingle Score-Control for Hallucination Reduction in Diffusion Models
Bhosale, Mahesh
Devulapally, Naresh Kumar
Wasi, Abdul
Pham, Chau
Lokhande, Vishnu Suresh
Doermann, David
Computer Vision and Pattern Recognition
Diffusion models have emerged as the backbone of modern generative AI, powering advances in vision, language, audio and other modalities. Despite their success, they suffer from hallucinations, implausible samples that lie outside the support of true data distribution, which degrade reliability and trust. In this work, we first empirically confirm previously proposed hypothesis that score smoothness causes hallucinations in Image Generation diffusion models and provide a density-based perspective. We further formalize this notion by linking the hallucinations probability mass to lipschitz constant of the learned score function. Motivated by this, we introduce a Variance-Guided Score Modulation (VSM) strategy that controls the score Jacobian, in turn reducing score smoothness and better approximating the ground truth score that decreases hallucinations. Empirical results on synthetic and real-world datasets demonstrate that our approach reduces hallucinations (up to ~25%) while maintaining high fidelity and diversity, providing a principled step toward more reliable diffusion-based image generation. We also propose two benchmark datasets with extreme semantic variation for systematic hallucination evaluation. Code and Datasets are publicly available at https://github.com/bhosalems/VSM.
title Score-Control for Hallucination Reduction in Diffusion Models
topic Computer Vision and Pattern Recognition
url https://arxiv.org/abs/2606.00377