Cover Image

Slowing Learning by Erasing Simple Features

Fuente: arXiv

Saved in:

Bibliographic Details
Main Authors:	Quirke, Lucia, Belrose, Nora
Format:	Preprint
Published:	2025
Subjects:	Machine Learning
Online Access:	Acceder al recurso
Tags:	Add Tag No Tags, Be the first to tag this record!

Similar Items

Binary Sparse Coding for Interpretability
by: Quirke, Lucia, et al.
Published: (2025)

Neural Networks Learn Statistics of Increasing Complexity
by: Belrose, Nora, et al.
Published: (2024)

Sparse Autoencoders Trained on the Same Data Learn Different Features
by: Paulo, Gonçalo, et al.
Published: (2025)

Estimating the Probability of Sampling a Trained Neural Network at Random
by: Scherlis, Adam, et al.
Published: (2025)

Evaluating SAE interpretability without explanations
by: Paulo, Gonçalo, et al.
Published: (2025)

Converting MLPs into Polynomials in Closed Form
by: Belrose, Nora, et al.
Published: (2025)

Balancing Label Quantity and Quality for Scalable Elicitation
by: Mallen, Alex, et al.
Published: (2024)

Understanding Gradient Descent through the Training Jacobian
by: Belrose, Nora, et al.
Published: (2024)

Partially Rewriting a Transformer in Natural Language
by: Paulo, Gonçalo, et al.
Published: (2025)

Examining Two Hop Reasoning Through Information Content Scaling
by: Johnston, David, et al.
Published: (2025)

Automatically Interpreting Millions of Features in Large Language Models
by: Paulo, Gonçalo, et al.
Published: (2024)

Transcoders Beat Sparse Autoencoders for Interpretability
by: Paulo, Gonçalo, et al.
Published: (2025)

Refusal in LLMs is an Affine Function
by: Marshall, Thomas, et al.
Published: (2024)

Does Transformer Interpretability Transfer to RNNs?
by: Paulo, Gonçalo, et al.
Published: (2024)

Mechanistic Anomaly Detection for "Quirky" Language Models
by: Johnston, David O., et al.
Published: (2025)

Eliciting Latent Knowledge from Quirky Language Models
by: Mallen, Alex, et al.
Published: (2023)

Understanding Addition in Transformers
by: Quirke, Philip, et al.
Published: (2023)

Learning Successor Features the Simple Way
by: Chua, Raymond, et al.
Published: (2024)

Understanding Addition and Subtraction in Transformers
by: Quirke, Philip, et al.
Published: (2024)

Slow Feature Analysis as Variational Inference Objective
by: Schüler, Merlin, et al.
Published: (2025)

Erasing the Bias: Fine-Tuning Foundation Models for Semi-Supervised Learning
by: Gan, Kai, et al.
Published: (2024)

Balancing Plasticity and Stability with Fast and Slow Successor Features
by: Chua, Raymond, et al.
Published: (2026)

LEACE: Perfect linear concept erasure in closed form
by: Belrose, Nora, et al.
Published: (2023)

Eliciting Latent Predictions from Transformers with the Tuned Lens
by: Belrose, Nora, et al.
Published: (2023)

What is the relation between Slow Feature Analysis and the Successor Representation?
by: Seabrook, Eddie, et al.
Published: (2024)

Slow Feature Analysis on Markov Chains from Goal-Directed Behavior
by: Schüler, Merlin, et al.
Published: (2025)

Erase or Hide? Suppressing Spurious Unlearning Neurons for Robust Unlearning
by: Yang, Nakyeong, et al.
Published: (2025)

Erasing Conceptual Knowledge from Language Models
by: Gandikota, Rohit, et al.
Published: (2024)

A Simple Data Augmentation for Feature Distribution Skewed Federated Learning
by: Yan, Yunlu, et al.
Published: (2023)

When Are Concepts Erased From Diffusion Models?
by: Lu, Kevin, et al.
Published: (2025)

Erase at the Core: Representation Unlearning for Machine Unlearning
by: Lee, Jaewon, et al.
Published: (2026)

Redirection for Erasing Memory (REM): Towards a universal unlearning method for corrupted data
by: Schoepf, Stefan, et al.
Published: (2025)

Flip Learning: Weakly Supervised Erase to Segment Nodules in Breast Ultrasound
by: Huang, Yuhao, et al.
Published: (2025)

Dichotomy of Feature Learning and Unlearning: Fast-Slow Analysis on Neural Networks with Stochastic Gradient Descent
by: Imai, Shota, et al.
Published: (2026)

EraseFlow: Learning Concept Erasure Policies via GFlowNet-Driven Alignment
by: Kusumba, Abhiram, et al.
Published: (2025)

Erasing Without Remembering: Implicit Knowledge Forgetting in Large Language Models
by: Wang, Huazheng, et al.
Published: (2025)

Side Effects of Erasing Concepts from Diffusion Models
by: Saha, Shaswati, et al.
Published: (2025)

Erased but Not Forgotten: How Backdoors Compromise Concept Erasure
by: Braun, Tobias, et al.
Published: (2025)

Erasing Undesirable Concepts in Diffusion Models with Adversarial Preservation
by: Bui, Anh, et al.
Published: (2024)

Simulating, Fast and Slow: Learning Policies for Black-Box Optimization
by: Massoli, Fabio Valerio, et al.
Published: (2024)