paper-with-me

Papers

Exploring Local Memorization in Diffusion Models via Bright Ending Attention

2024-10-29 · Chen Chen, Daochang Liu, Mubarak Shah, Chang Xu

In this paper, we identify and leverage a novel `bright ending' (BE) anomaly in diffusion models prone to memorizing training images to address a new task: locating localized memorization regions within these models. BE refers to a distinct cross-attention pattern observed in text-to-image generations using diffusion models. Specifically, memorized image patches exhibit significantly greater attention to the end token during the final inference step compared to non-memorized patches. This attention map effectively highlights regions where the generated image replicates training data. Furthermore, driven by our observation that local memorization significantly underperforms in existing tasks of measuring, detecting, and mitigating memorization in diffusion models compared to global memorization, we propose a simple yet effective method to integrate BE and the results of the new localization task into these existing frameworks. This integration effectively improves their performances by narrowing the performance gap caused by local memorization. Our results not only demonstrate the successful execution of the new localization task but also establish new state-of-the-art performance across all existing tasks, underscoring the significance of the BE phenomenon.

📄 PDF Abstract BibTeX arXiv:2410.21665

Code (0)

등록된 구현이 없습니다.

Tasks

Memorization

Methods 이 논문이 사용한 방법론

Softmax The Softmax output function transforms a previous layer's output into a vector of probabilities. It is commonly used for multiclass classification. Given an input vector $x$…
Attention 설명 없음
Diffusion Diffusion models generate samples by gradually removing noise from a signal, and their training objective can be expressed as a reweighted variational lower-bound…

Similar Papers 제목 키워드 기반

Local Coverage Governs Memorization in Diffusion Models

2026-06-12 · Claudia Merger, Sebastian Goldt arxiv

Memorization in diffusion models is often treated as a global property of the model or dataset. In practice, however, a single diffusion model can simultaneously generate both memorized and novel samples. Which training …

Density Estimation

Mitigating data replication in text-to-audio generative diffusion models through anti-memorization guidance

2025-09-18 · Francisco Messina, Francesca Ronchini, Luca Comanducci, Paolo Bestagini 외 arxiv

A persistent challenge in generative audio models is data replication, where the model unintentionally generates parts of its training data during inference. In this work, we address this issue in text-to-audio diffusion…

Audio Generation

Losing dimensions: Geometric memorization in generative diffusion

2024-10-11 · Beatrice Achilli, Enrico Ventura, Gianluigi Silvestri, Bao Pham 외

Generative diffusion processes are state-of-the-art machine learning models deeply connected with fundamental concepts in statistical physics. Depending on the dataset size and the capacity of the network, their behavior…

Memorization

Localizing Memorized Regions in Diffusion Models via Coordinate-Wise Curvature Differences

2026-05-26 · Gwangho Kim, Sungyoon Lee arxiv

Diffusion models can unintentionally memorize training samples, raising concerns about privacy and copyright. While recent methods can detect memorization, they often rely on global or model-specific signals and provide …

Latent Diffusion Inversion Requires Understanding the Latent Space

2025-11-25 · Mingxing Rao, Bowen Qu, Daniel Moyer arxiv

The recovery of training data from generative models ("model inversion") has been extensively studied for diffusion models in the data domain as a memorization/overfitting phenomenon. Latent diffusion models (LDMs), whic…