Image de couverture de livre

Sparse-Autoencoder-Guided Internal Representation Unlearning for Large Language Models

Fuente: arXiv

Enregistré dans:

Détails bibliographiques
Auteurs principaux:	Yamashita, Tomoya, Ito, Akira, Yamanaka, Yuuki, Yamada, Masanori, Miura, Takayuki, Shibahara, Toshiki
Format:	Preprint
Publié:	2025
Sujets:	Computation and Language Machine Learning
Accès en ligne:	Acceder al recurso
Tags:	Ajouter un tag Pas de tags, Soyez le premier à ajouter un tag!

Documents similaires

Concept Unlearning in Large Language Models via Self-Constructed Knowledge Triplets
par: Yamashita, Tomoya, et autres
Publié: (2025)

Robust Backdoor Removal by Reconstructing Trigger-Activated Changes in Latent Representation
par: Iwahana, Kazuki, et autres
Publié: (2025)

One-Shot Machine Unlearning with Mnemonic Code
par: Yamashita, Tomoya, et autres
Publié: (2023)

Model Unlearning via Sparse Autoencoder Subspace Guided Projections
par: Wang, Xu, et autres
Publié: (2025)

A Survey on Sparse Autoencoders: Interpreting the Internal Mechanisms of Large Language Models
par: Shu, Dong, et autres
Publié: (2025)

How a Bilingual LM Becomes Bilingual: Tracing Internal Representations with Sparse Autoencoders
par: Inaba, Tatsuro, et autres
Publié: (2025)

Guiding LLM Post-training Data Engineering with Model Internals from Sparse Autoencoders
par: Jing, Yi, et autres
Publié: (2026)

The Geometry of Tokens in Internal Representations of Large Language Models
par: Viswanathan, Karthik, et autres
Publié: (2025)

Rethinking Machine Unlearning for Large Language Models
par: Liu, Sijia, et autres
Publié: (2024)

Sparse Shift Autoencoders for Identifying Concepts from Large Language Model Activations
par: Joshi, Shruti, et autres
Publié: (2025)

Sparse Autoencoders Enable Scalable and Reliable Circuit Identification in Language Models
par: O'Neill, Charles, et autres
Publié: (2024)

SAEBench: A Comprehensive Benchmark for Sparse Autoencoders in Language Model Interpretability
par: Karvonen, Adam, et autres
Publié: (2025)

SAEs $\textit{Can}$ Improve Unlearning: Dynamic Sparse Autoencoder Guardrails for Precision Unlearning in LLMs
par: Muhamed, Aashiq, et autres
Publié: (2025)

Large Language Model Unlearning
par: Yao, Yuanshun, et autres
Publié: (2023)

Positive-Unlabeled Diffusion Models for Preventing Sensitive Data Generation
par: Takahashi, Hiroshi, et autres
Publié: (2025)

Dissecting Fine-Tuning Unlearning in Large Language Models
par: Hong, Yihuai, et autres
Publié: (2024)

GUARD: Guided Unlearning and Retention via Data Attribution for Large Language Models
par: Niu, Peizhi, et autres
Publié: (2025)

Truth as a Trajectory: What Internal Representations Reveal About Large Language Model Reasoning
par: Damirchi, Hamed, et autres
Publié: (2026)

Quantifying Feature Space Universality Across Large Language Models via Sparse Autoencoders
par: Lan, Michael, et autres
Publié: (2024)

Training Superior Sparse Autoencoders for Instruct Models
par: Li, Jiaming, et autres
Publié: (2025)

Evaluating Adversarial Robustness of Concept Representations in Sparse Autoencoders
par: Li, Aaron J., et autres
Publié: (2025)

Offset Unlearning for Large Language Models
par: Huang, James Y., et autres
Publié: (2024)

ZeroUnlearn: Few-Shot Knowledge Unlearning in Large Language Models
par: Lin, Yujie, et autres
Publié: (2026)

Is the Hard-Label Cryptanalytic Model Extraction Really Polynomial?
par: Ito, Akira, et autres
Publié: (2025)

Rethinking Evaluation of Sparse Autoencoders through the Representation of Polysemous Words
par: Minegishi, Gouki, et autres
Publié: (2025)

Alternate Preference Optimization for Unlearning Factual Knowledge in Large Language Models
par: Mekala, Anmol, et autres
Publié: (2024)

Hierarchical Federated Unlearning for Large Language Models
par: Zhong, Yisheng, et autres
Publié: (2025)

Soft Prompting for Unlearning in Large Language Models
par: Bhaila, Karuna, et autres
Publié: (2024)

Multi-Objective Large Language Model Unlearning
par: Pan, Zibin, et autres
Publié: (2024)

Model Directions, Not Words: Mechanistic Topic Models Using Sparse Autoencoders
par: Zheng, Carolina, et autres
Publié: (2025)

SparseEval: Efficient Evaluation of Large Language Models by Sparse Optimization
par: Zhang, Taolin, et autres
Publié: (2026)

Feature Rivalry in Sparse Autoencoder Representations: A Mechanistic Study of Uncertainty-Driven Feature Competition in LLMs
par: Harshavardhan
Publié: (2026)

Existing Large Language Model Unlearning Evaluations Are Inconclusive
par: Feng, Zhili, et autres
Publié: (2025)

Q-Sparse: All Large Language Models can be Fully Sparsely-Activated
par: Wang, Hongyu, et autres
Publié: (2024)

AlignSAE: Concept-Aligned Sparse Autoencoders
par: Yang, Minglai, et autres
Publié: (2025)

DLM-Scope: Mechanistic Interpretability of Diffusion Language Models via Sparse Autoencoders
par: Wang, Xu, et autres
Publié: (2026)

Diversity-driven Data Selection for Language Model Tuning through Sparse Autoencoder
par: Yang, Xianjun, et autres
Publié: (2025)

Attention Sinks as Internal Signals for Hallucination Detection in Large Language Models
par: Binkowski, Jakub, et autres
Publié: (2026)

Necessary and Sufficient Watermark for Large Language Models
par: Takezawa, Yuki, et autres
Publié: (2023)

Sparse Autoencoder Decomposition of Clinical Sequence Model Representations: Feature Complexity, Task Specialisation, and Mortality Prediction
par: Sainsbury, Chris, et autres
Publié: (2026)