paper-with-me

홈 › Papers

Capacity Control is an Effective Memorization Mitigation Mechanism in Text-Conditional Diffusion Models

2024-10-29 · Raman Dutt, Pedro Sanchez, Ondrej Bohdal, Sotirios A. Tsaftaris, Timothy Hospedales

In this work, we present compelling evidence that controlling model capacity during fine-tuning can effectively mitigate memorization in diffusion models. Specifically, we demonstrate that adopting Parameter-Efficient Fine-Tuning (PEFT) within the pre-train fine-tune paradigm significantly reduces memorization compared to traditional full fine-tuning approaches. Our experiments utilize the MIMIC dataset, which comprises image-text pairs of chest X-rays and their corresponding reports. The results, evaluated through a range of memorization and generation quality metrics, indicate that PEFT not only diminishes memorization but also enhances downstream generation quality. Additionally, PEFT methods can be seamlessly combined with existing memorization mitigation techniques for further improvement. The code for our experiments is available at: https://github.com/Raman1121/Diffusion_Memorization_HPO

📄 PDF Abstract BibTeX arXiv:2410.22149

Code (1)

raman1121/diffusion_memorization_hpo 공식 구현 pytorch

Tasks

Memorizationparameter-efficient fine-tuning

Methods 이 논문이 사용한 방법론

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

MemControl: Mitigating Memorization in Diffusion Models via Automated Parameter Selection

2024-05-29 · Raman Dutt, Ondrej Bohdal, Pedro Sanchez, Sotirios A. Tsaftaris 외

Diffusion models excel in generating images that closely resemble their training data but are also susceptible to data memorization, raising privacy, ethical, and legal concerns, particularly in sensitive domains such as…

Image GenerationMedical Image GenerationMemorizationparameter-efficient fine-tuning

The Landscape of Memorization in LLMs: Mechanisms, Measurement, and Mitigation

2025-07-08 · Alexander Xiong, Xuandong Zhao, Aneesh Pappu, Dawn Song arxiv

Large Language Models (LLMs) have demonstrated remarkable capabilities across a wide range of tasks, yet they also exhibit memorization of their training data. This phenomenon raises critical questions about model behavi…

Exploring Memorization in Adversarial Training

2021-06-03 · ICLR 2022 4 · Yinpeng Dong, Ke Xu, Xiao Yang, Tianyu Pang 외

Deep learning models have a propensity for fitting the entire training set even with random labels, which requires memorization of every training sample. In this paper, we explore the memorization effect in adversarial t…

Memorization

GSS: Gated Subspace Steering for Selective Memorization Mitigation in LLMs

2026-02-09 · Xuanqi Zhang, Haoyang Shang, Xiaoxiao Li arxiv

Large language models (LLMs) can memorize and reproduce training sequences verbatim -- a tendency that undermines both generalization and privacy. Existing mitigation methods apply interventions uniformly, degrading perf…

Demystifying Foreground-Background Memorization in Diffusion Models

2025-08-16 · Jimmy Z. Di, Yiwei Lu, Yaoliang Yu, Gautam Kamath 외 arxiv

Diffusion models (DMs) memorize training images and can reproduce near-duplicates during generation. Current detection methods identify verbatim memorization but fail to capture two critical aspects: quantifying partial …