Geometric Erasure by Contrastive Velocity Matching in Rectified Flows

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Grebe, Jonas Henry, Braun, Tobias, Rohrbach, Anna, Rohrbach, Marcus
Format: Preprint
Published: 2026
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866914619416641536
author Grebe, Jonas Henry
Braun, Tobias
Rohrbach, Anna
Rohrbach, Marcus
author_facet Grebe, Jonas Henry
Braun, Tobias
Rohrbach, Anna
Rohrbach, Marcus
contents While the rapid adoption of multimodal generative models offers immense potential, it has also increased the risks of harmful content synthesis, deepfakes, and copyright infringements. To address these challenges, concept erasure has emerged as a prospective safeguard. However, as the field gradually transitions from U-Net-based diffusion models to Rectified Flow Transformers, erasure research has struggled to keep pace. In this work, we introduce GEM, a simple but highly effective erasure framework for Rectified Flow models. As part of our contribution, we establish a principled bridge between trajectory-based unlearning grounded in Generative Flow Networks and classic teacher-guided erasure: we translate trajectory-based signals into a teacher-guided flow-matching setup that unifies the strengths of both paradigms. Concretely, a teacher provides complementary attraction and repulsion signals that we combine into a single geometric guidance objective, yielding targeted suppression of unwanted concepts while preserving benign generation.
format Preprint
id arxiv_https___arxiv_org_abs_2606_00140
institution arXiv
publishDate 2026
record_format arxiv
spellingShingle Geometric Erasure by Contrastive Velocity Matching in Rectified Flows
Grebe, Jonas Henry
Braun, Tobias
Rohrbach, Anna
Rohrbach, Marcus
Machine Learning
Artificial Intelligence
While the rapid adoption of multimodal generative models offers immense potential, it has also increased the risks of harmful content synthesis, deepfakes, and copyright infringements. To address these challenges, concept erasure has emerged as a prospective safeguard. However, as the field gradually transitions from U-Net-based diffusion models to Rectified Flow Transformers, erasure research has struggled to keep pace. In this work, we introduce GEM, a simple but highly effective erasure framework for Rectified Flow models. As part of our contribution, we establish a principled bridge between trajectory-based unlearning grounded in Generative Flow Networks and classic teacher-guided erasure: we translate trajectory-based signals into a teacher-guided flow-matching setup that unifies the strengths of both paradigms. Concretely, a teacher provides complementary attraction and repulsion signals that we combine into a single geometric guidance objective, yielding targeted suppression of unwanted concepts while preserving benign generation.
title Geometric Erasure by Contrastive Velocity Matching in Rectified Flows
topic Machine Learning
Artificial Intelligence
url https://arxiv.org/abs/2606.00140