FlowErase-RL applies GRPO to reformulate concept erasure in flow matching models as reward optimization using a dynamic dual-path mechanism for target suppression and non-target preservation.
Saeuron: Interpretable concept unlearning in diffusion models with sparse autoencoders
9 Pith papers cite this work. Polarity classification is still indexing.
9
Pith papers citing it
citation-role summary
background 3
citation-polarity summary
verdicts
UNVERDICTED 9roles
background 3polarities
background 3representative citing papers
LVO applies optimization-based feature visualization to latent diffusion models after disentangling their representations with sparse autoencoders, yielding recognizable concept images on a fine-tuned Stable Diffusion model that are clearer than those from entangled baselines.
A cross-attention SAE with sparsemax attention achieves lower reconstruction loss and higher-quality concepts than fixed-sparsity baselines by making activation counts data-dependent.
SafeDiffusion-R1 uses online GRPO with CLIP embedding steering to cut inappropriate content from 48.9% to 18.07% and nudity detections from 646 to 15 in diffusion models while raising GenEval scores from 42.08% to 47.83% and generalizing across seven harm categories without supervised pairs or extra
SoftSAE replaces fixed-K sparsity in autoencoders with a learned, input-dependent number of active features via a soft top-k operator.
A training-free double-projection linear transformation erases target concepts from generative models by computing a proxy projection then applying a constrained update in the left null space of known directions.
BARRIER applies interval arithmetic to SVD-based activation projections to create bounded forget regions that enable aggressive unlearning while providing formal protection for retain distributions via tail bounds on functional drift.
SPACE induces sparsity in cross-attention parameters via closed-form iterative updates to erase target concepts more effectively than dense baselines in large diffusion models.
A survey classifying machine unlearning into centralized (exact and approximate), distributed/irregular data, verification, and privacy/security categories with technique overviews.
citing papers explorer
-
FlowErase-RL: Rethinking Concept Erasure as Reward Optimization in Flow Matching Models
FlowErase-RL applies GRPO to reformulate concept erasure in flow matching models as reward optimization using a dynamic dual-path mechanism for target suppression and non-target preservation.
-
Deep Dreams Are Made of This: Visualizing Monosemantic Features in Diffusion Models
LVO applies optimization-based feature visualization to latent diffusion models after disentangling their representations with sparse autoencoders, yielding recognizable concept images on a fine-tuned Stable Diffusion model that are clearer than those from entangled baselines.
-
Improving Sparse Autoencoder with Dynamic Attention
A cross-attention SAE with sparsemax attention achieves lower reconstruction loss and higher-quality concepts than fixed-sparsity baselines by making activation counts data-dependent.
-
SafeDiffusion-R1: Online Reward Steering for Safe Diffusion Post-Training
SafeDiffusion-R1 uses online GRPO with CLIP embedding steering to cut inappropriate content from 48.9% to 18.07% and nudity detections from 646 to 15 in diffusion models while raising GenEval scores from 42.08% to 47.83% and generalizing across seven harm categories without supervised pairs or extra
-
SoftSAE: Dynamic Top-K Selection for Adaptive Sparse Autoencoders
SoftSAE replaces fixed-K sparsity in autoencoders with a learned, input-dependent number of active features via a soft top-k operator.
-
Closed-Form Concept Erasure via Double Projections
A training-free double-projection linear transformation erases target concepts from generative models by computing a proxy projection then applying a constrained update in the left null space of known directions.
-
BARRIER: Bounded Activation Regions for Robust Information Erasure
BARRIER applies interval arithmetic to SVD-based activation projections to create bounded forget regions that enable aggressive unlearning while providing formal protection for retain distributions via tail bounds on functional drift.
-
Empty SPACE: Cross-Attention Sparsity for Concept Erasure in Diffusion Models
SPACE induces sparsity in cross-attention parameters via closed-form iterative updates to erase target concepts more effectively than dense baselines in large diffusion models.
-
Machine Unlearning: A Comprehensive Survey
A survey classifying machine unlearning into centralized (exact and approximate), distributed/irregular data, verification, and privacy/security categories with technique overviews.