---
title: "Difficulty Calibration Across Time for Membership Inference in Diffusion Models"
source: "arXiv AI/CL/LG（论文）"
source_tier: "t2"
url: "https://arxiv.org/abs/2610.40616v1"
permalink: "https://feed.astralor.com/items/itm_208f2a6d1903c57b4caef5dc"
published: "2026-10-29T17:54:54.000Z"
score: 57
channel: "tech"
tags: ["推理效率","Difficulty Calibration Across","Time","Membership Inference","Diffusion Models Modern","Membership","MIAs","Modern"]
---

# Difficulty Calibration Across Time for Membership Inference in Diffusion Models

## 速览

Modern diffusion models are trained on massive datasets that often include copyrighted or private images. Membership inference attacks (MIAs) address this by testing whether a given image was in a model's training set. …

## 关键点

- Modern diffusion models are trained on massive datasets that often include copyrighted or private images.
- Membership inference attacks (MIAs) address this by testing whether a given image was in a model's training set.
- The standard approach scores an image by its reconstruction loss and flags low-loss images as members.

## 全文

Modern diffusion models are trained on massive datasets that often include copyrighted or private images. Membership inference attacks (MIAs) address this by testing whether a given image was in a model's training set. The standard approach scores an image by its reconstruction loss and flags low-loss images as members. This often fails because some images are simply easier to reconstruct than others, so a simple unseen image can score a lower loss than a complex training image. Difficulty calibration separates memorization from how difficult an image is to begin with. We show how to calibrate along two axes of time: training time and diffusion time. First, across training time, we introduce Reference-Model Calibration (RMC), which compares a fine-tuned model's loss against an earlier checkpoint. This cleanly isolates newly memorized data and gives near-perfect auditing, even without text captions. Since earlier checkpoints are not always available, we next present Temporal Self-Calibration (TSC), which calibrates across diffusion time using only a single model. Our key observation is that an image's difficulty persists across noise levels, whereas memorization is localized in time. By comparing the loss at a target timestep against neighboring timesteps from the same model, TSC attenuates image difficulty using the model's own trajectory. Empirically, our methods beat the state of the art by a wide margin, and unlike most existing attacks, generalize across different model parameterizations such as $x$-, $v$-, and $ε$-prediction models. Code is available at https://github.com/MichaelJoseph777/TSC.

---

引用前请核对原文;本文由星际流动整理:https://feed.astralor.com/items/itm_208f2a6d1903c57b4caef5dc