The paper introduces SCALPEL, a contrastive sparse autoencoder that selectively learns to unlearn target-specific information from neural network representations. It addresses the energy bias in standard reconstruction-based extractors, which tend to preserve dominant background structure while neglecting low-energy target components. The authors provide theoretical justification that contrastive training enhances target selectivity and demonstrate experimentally that SCALPEL outperforms NMF and standard SAE interventions on the TOFU benchmark, achieving performance comparable to Gradient Difference and RMU methods.
arXiv:2609.16229v1 Announce Type: new Abstract: Machine unlearning aims to remove specific knowledge from a trained large language model (LLM) without retraining from scratch. Existing methods modify...
The paper introduces Dynamic DAE Guardrails (DSG), a method that uses Dynamic Sparse Autoencoders to perform precision unlearning in large language models. DSG leverages principled feature selection and a dynamic classifier to target activation-based unlearning, outperforming existing gradient‑based methods in terms of computational efficiency, stability, sequential unlearning, resistance to relearning attacks, data efficiency, and interpretability.
The paper introduces ADU, a fine‑grained training framework that unlearns sensitive information from large language models by decoupling contextual attention pathways instead of erasing tokens. ADU exploits the distinction between local and global attention heads to identify and suppress attention paths that retrieve persistent sensitive anchors, while preserving local‑attention structure and overall language modeling performance. Evaluation on the TOFU and WMDP benchmarks shows ADU achieves superior forget quality (0.93 on TOFU) and retains 92.9% of model utility compared to 81.9% for existing baselines, with fewer side effects in benign contexts.
Machine unlearning aims to remove the influence of specific training data while preserving model utility. Many state-of-the-art approaches pursue this goal by restricting the forgetting update to a subset of parameters selected through gradient-based saliency.
arXiv:2606. 26620v1 Announce Type: cross Abstract: Sparse autoencoders (SAEs) have emerged as a powerful tool for decomposing superposed language model representations into sparse and interpretable features.
arXiv:2511. 20196v2 Announce Type: replace Abstract: Multimodal large language models (MLLMs) can inadvertently memorize privacy-sensitive information during training.
arXiv:2606. 14990v1 Announce Type: cross Abstract: Sparse autoencoders (SAEs) are standard tools for mechanistic interpretability, but current SAE families are constrained by fixed encoder nonlinearities such as ReLU, JumpReLU, and TopK.
arXiv:2606. 30609v1 Announce Type: cross Abstract: Sparse Autoencoders (SAEs) are widely used to interpret large language models by decomposing activations into sparse, human-understandable features, but scaling to large dictionaries exposes fundamental challenges.
arXiv:2606. 11722v1 Announce Type: cross Abstract: Finding interpretable directions in language-model representations is critical for understanding and controlling model behavior.
arXiv:2606. 14040v1 Announce Type: new Abstract: Sparse autoencoders (SAEs) are typically trained to reconstruct the \textbf{entire} residual stream through a sparse dictionary, implicitly assuming that all activation content is amenable to sparse, monosemantic decomposition.
arXiv:2606. 06320v1 Announce Type: new Abstract: Machine unlearning aims to remove targeted knowledge from a trained model while preserving its general capabilities.
arXiv:2606. 31699v1 Announce Type: cross Abstract: Sparse autoencoders (SAEs) have recently been proposed as interpretable tools for concept-level manipulation, under the assumption that isolated features can serve as controllable intervention points.