REVIEW 8 cited by
SAeUron: Interpretable Concept Unlearning in Diffusion Models with Sparse Autoencoders
Not yet reviewed by Pith; the record is open.
This paper has not been read by Pith yet. Machine review is queued; the pith claim, tier, and objections will appear here once it completes.
SPECIMEN: schema-true, not a live event
T0 review · schema-true
One-sentence machine reading of the paper's core claim.
pith:XXXXXXXX · record.json · timestamp
read the original abstract
Diffusion models, while powerful, can inadvertently generate harmful or undesirable content, raising significant ethical and safety concerns. Recent machine unlearning approaches offer potential solutions but often lack transparency, making it difficult to understand the changes they introduce to the base model. In this work, we introduce SAeUron, a novel method leveraging features learned by sparse autoencoders (SAEs) to remove unwanted concepts in text-to-image diffusion models. First, we demonstrate that SAEs, trained in an unsupervised manner on activations from multiple denoising timesteps of the diffusion model, capture sparse and interpretable features corresponding to specific concepts. Building on this, we propose a feature selection method that enables precise interventions on model activations to block targeted content while preserving overall performance. Our evaluation shows that SAeUron outperforms existing approaches on the UnlearnCanvas benchmark for concepts and style unlearning, and effectively eliminates nudity when evaluated with I2P. Moreover, we show that with a single SAE, we can remove multiple concepts simultaneously and that in contrast to other methods, SAeUron mitigates the possibility of generating unwanted content under adversarial attack. Code and checkpoints are available at https://github.com/cywinski/SAeUron.
Forward citations
Cited by 8 Pith papers
-
You Don't Need All That Attention: Surgical Memorization Mitigation in Text-to-Image Diffusion Models
A per-prompt cross-attention spike detector plus repulsive-attractive guidance (GUARD) substantially reduces verbatim and template memorization in Stable Diffusion at inference time.
-
Measuring Monosemanticity in Sparse Autoencoders via Latent Activation Coherence
A label-free monosemanticity metric for sparse autoencoders — average pairwise Jaccard similarity of binarized activation patterns — is more robust to embedding anisotropy than cosine-based MS.
-
UnHype: CLIP-Guided Hypernetworks for Dynamic LoRA Unlearning
UnHype generates concept-specific LoRA unlearning weights on the fly from CLIP text embeddings by training a hypernetwork to follow the gradient of an unlearning loss, enabling single- and multi-concept erasure in dif...
-
The SuperActivator Mechanism: Transformers Concentrate Reliable Concept Signals in the Tail
Reliable concept presence in transformers is concentrated in the extreme high-activation tail of in-concept tokens; thresholding that tail improves concept detection and localization.
-
FoldSAE: Learning to Steer Protein Folding Through Sparse Representations
Sparse-autoencoder features inside RFdiffusion predict—and, when amplified or suppressed, causally steer—helix and strand content in generated protein backbones.
-
SAEmnesia: Erasing Concepts in Diffusion Models with Supervised Sparse Autoencoders
A supervised sparse autoencoder binds each concept to a single neuron, letting Stable Diffusion erase a concept by steering one latent.
-
CytoSAE: Interpretable Cell Embeddings for Hematology
CytoSAE learns sparse, expert-validated morphological concepts from blood-cell images that generalize across datasets and can classify AML subtypes at patient level with F1 0.83.
-
Beyond Black Boxes: Enhancing Interpretability of Transformers Trained on Neural Data
Inserting a TopK sparse autoencoder into the POYO+ neural decoder preserves performance while yielding latent units selective for orientation, temporal frequency, and genetic background, and ablating them causally imp...
Discussion (0). Sign in to comment.