English
Related papers

Related papers: Linear Adversarial Concept Erasure

200 papers

Text-to-image generative models can produce photo-realistic images for an extremely broad range of concepts, and their usage has proliferated widely among the general public. On the flip side, these models have numerous drawbacks, including…

Machine Learning · Computer Science 2023-10-10 Minh Pham , Kelly O. Marshall , Niv Cohen , Govind Mittal , Chinmay Hegde

Machine unlearning is a key defense mechanism for removing unauthorized concepts from text-to-image diffusion models, yet recent evidence shows that latent visual information often persists after unlearning. Existing adversarial approaches…

Computer Vision and Pattern Recognition · Computer Science 2026-02-03 Ignacy Kolton , Kacper Marzol , Paweł Batorski , Marcin Mazur , Paul Swoboda , Przemysław Spurek

Generating images from text has become easier because of the scaling of diffusion models and advancements in the field of vision and language. These models are trained using vast amounts of data from the Internet. Hence, they often contain…

Computer Vision and Pattern Recognition · Computer Science 2024-08-30 Masane Fuchi , Tomohiro Takagi

Adversarial representation learning is a promising paradigm for obtaining data representations that are invariant to certain sensitive attributes while retaining the information necessary for predicting target attributes. Existing…

Machine Learning · Computer Science 2019-12-30 Bashir Sadeghi , Runyi Yu , Vishnu Naresh Boddeti

We investigate the behavior of methods that use linear projections to remove information about a concept from a language representation, and we consider the question of what happens to a dataset transformed by such a method. A theoretical…

Computation and Language · Computer Science 2024-03-26 Richard Johansson

Neural networks have greatly boosted performance in computer vision by learning powerful representations of input data. The drawback of end-to-end training for maximal overall performance are black-box models whose hidden representations…

Computer Vision and Pattern Recognition · Computer Science 2020-04-29 Patrick Esser , Robin Rombach , Björn Ommer

One of the main methods for computational interpretation of a text is mapping it into a vector in some embedding space. Such vectors can then be used for a variety of textual processing tasks. Recently, most embedding spaces are a product…

Computation and Language · Computer Science 2023-11-10 Adi Simhi , Shaul Markovitch

With widening deployments of natural language processing (NLP) in daily life, inherited social biases from NLP models have become more severe and problematic. Previous studies have shown that word embeddings trained on human-generated…

Computation and Language · Computer Science 2021-12-13 Lei Ding , Dengdeng Yu , Jinhan Xie , Wenxing Guo , Shenggang Hu , Meichen Liu , Linglong Kong , Hongsheng Dai , Yanchun Bao , Bei Jiang

Motivated by recent advancements in text-to-image diffusion, we study erasure of specific concepts from the model's weights. While Stable Diffusion has shown promise in producing explicit or realistic artwork, it has raised concerns…

Computer Vision and Pattern Recognition · Computer Science 2023-06-22 Rohit Gandikota , Joanna Materzynska , Jaden Fiotto-Kaufman , David Bau

Text-to-Image models such as Stable Diffusion have shown impressive image generation synthesis, thanks to the utilization of large-scale datasets. However, these datasets may contain sexually explicit, copyrighted, or undesirable content,…

Computer Vision and Pattern Recognition · Computer Science 2023-12-21 Seunghoo Hong , Juhun Lee , Simon S. Woo

Traditional neural embeddings represent concepts as points, excelling at similarity but struggling with higher-level reasoning and asymmetric relationships. We introduce a novel paradigm: embedding concepts as linear subspaces. This…

Machine Learning · Computer Science 2025-08-26 Gabriel Moreira , Zita Marinho , Manuel Marques , João Paulo Costeira , Chenyan Xiong

Enforcing complex (e.g., nonconvex) operational constraints is a critical challenge in real-world learning and control systems. However, existing methods struggle to efficiently enforce general classes of constraints. To address this, we…

Machine Learning · Computer Science 2026-04-07 Maria Chzhen , Priya L. Donti

To what extent does concept erasure eliminate generative capacity in diffusion models? While prior evaluations have primarily focused on measuring concept suppression under specific textual prompts, we explore a complementary and…

Computer Vision and Pattern Recognition · Computer Science 2025-09-19 Ping Liu , Chi Zhang

Fairness has become a central issue for our research community as classification algorithms are adopted in societally critical domains such as recidivism prediction and loan approval. In this work, we consider the potential bias based on…

Machine Learning · Computer Science 2019-05-01 Rui Feng , Yang Yang , Yuehan Lyu , Chenhao Tan , Yizhou Sun , Chunping Wang

Advances in model editing through neuron pruning hold promise for removing undesirable concepts from large language models. However, it remains unclear whether models have the capacity to reacquire pruned concepts after editing. To…

Artificial Intelligence · Computer Science 2024-01-04 Michelle Lo , Shay B. Cohen , Fazl Barez

Semantic segmentation is a task that traditionally requires a large dataset of pixel-level ground truth labels, which is time-consuming and expensive to obtain. Recent advancements in the weakly-supervised setting show that reasonable…

Computer Vision and Pattern Recognition · Computer Science 2020-11-10 Erik Stammes , Tom F. H. Runia , Michael Hofmann , Mohsen Ghafoorian

In this paper, we present a method for adversarial decomposition of text representation. This method can be used to decompose a representation of an input sentence into several independent vectors, each of them responsible for a specific…

Computation and Language · Computer Science 2019-04-11 Alexey Romanov , Anna Rumshisky , Anna Rogers , David Donahue

Concept erasure helps stop diffusion models (DMs) from generating harmful content; but current methods face robustness retention trade off. Robustness means the model fine-tuned by concept erasure methods resists reactivation of erased…

Machine Learning · Computer Science 2026-02-16 Fengpeng Li , Kemou Li , Qizhou Wang , Bo Han , Jiantao Zhou

The linear subspace hypothesis (Bolukbasi et al., 2016) states that, in a language model's representation space, all information about a concept such as verbal number is encoded in a linear subspace. Prior work has relied on auxiliary…

Computation and Language · Computer Science 2025-03-27 Clément Guerner , Tianyu Liu , Anej Svete , Alexander Warstadt , Ryan Cotterell

The remarkable development of text-to-image generation models has raised notable security concerns, such as the infringement of portrait rights and the generation of inappropriate content. Concept erasure has been proposed to remove the…

Computer Vision and Pattern Recognition · Computer Science 2024-10-15 Yufan Liu , Jinyang An , Wanqian Zhang , Ming Li , Dayan Wu , Jingzi Gu , Zheng Lin , Weiping Wang