Cover Image

SAGE: Scalable Ground Truth Evaluations for Large Sparse Autoencoders

Fuente: arXiv

Saved in:

Bibliographic Details
Main Authors:	Venhoff, Constantin, Calinescu, Anisoara, Torr, Philip, de Witt, Christian Schroeder
Format:	Preprint
Published:	2024
Subjects:	Machine Learning
Online Access:	Acceder al recurso
Tags:	Add Tag No Tags, Be the first to tag this record!

Similar Items

Towards Understanding Multimodal Fine-Tuning: Spatial Features
by: Naghashyar, Lachin, et al.
Published: (2026)

Mixture of Experts Made Intrinsically Interpretable
by: Yang, Xingyi, et al.
Published: (2025)

DEEDEE: Fast and Scalable Out-of-Distribution Dynamics Detection
by: Aljaafari, Tala, et al.
Published: (2025)

Too Late to Recall: Explaining the Two-Hop Problem in Multimodal Knowledge Retrieval
by: Venhoff, Constantin, et al.
Published: (2025)

Understanding Reasoning in Thinking Language Models via Steering Vectors
by: Venhoff, Constantin, et al.
Published: (2025)

Base Models Know How to Reason, Thinking Models Learn When
by: Venhoff, Constantin, et al.
Published: (2025)

Efficient Dictionary Learning with Switch Sparse Autoencoders
by: Mudide, Anish, et al.
Published: (2024)

How Visual Representations Map to Language Feature Space in Multimodal LLMs
by: Venhoff, Constantin, et al.
Published: (2025)

Do Sparse Autoencoders Generalize? A Case Study of Answerability
by: Heindrich, Lovis, et al.
Published: (2025)

Toward Robust Real-World Audio Deepfake Detection: Closing the Explainability Gap
by: Channing, Georgia, et al.
Published: (2024)

Using causal abstractions to accelerate decision-making in complex bandit problems
by: Dyer, Joel, et al.
Published: (2025)

Reasoning-Finetuning Repurposes Latent Representations in Base Models
by: Ward, Jake, et al.
Published: (2025)

Quantifying Feature Space Universality Across Large Language Models via Sparse Autoencoders
by: Lan, Michael, et al.
Published: (2024)

Neural Network-Based Parameter Estimation of a Labour Market Agent-Based Model
by: Alves, M Lopes, et al.
Published: (2026)

Sandbagging in a Simple Survival Bandit Problem
by: Dyer, Joel, et al.
Published: (2025)

PSyDUCK: Training-Free Steganography for Latent Diffusion
by: Mahfuz, Aqib, et al.
Published: (2025)

Detecting Multi-Agent Collusion Through Multi-Agent Interpretability
by: Rose, Aaron, et al.
Published: (2026)

Bayesian Decision Making around Experts
by: Ornia, Daniel Jarne, et al.
Published: (2025)

Sensitivity Analysis for Climate Science with Generative Flow Models
by: Dobra, Alex, et al.
Published: (2025)

SynthSAEBench: Evaluating Sparse Autoencoders on Scalable Realistic Synthetic Data
by: Chanin, David, et al.
Published: (2026)

h1: Bootstrapping LLMs to Reason over Longer Horizons via Reinforcement Learning
by: Motwani, Sumeet Ramesh, et al.
Published: (2025)

IMAGIC-500: IMputation benchmark on A Generative Imaginary Country (500k samples)
by: Sun, Siyi, et al.
Published: (2025)

Evaluating Sparse Autoencoders for Monosemantic Representation
by: Fereidouni, Moghis, et al.
Published: (2025)

Causally Abstracted Multi-armed Bandits
by: Zennaro, Fabio Massimo, et al.
Published: (2024)

Rethinking Out-of-Distribution Detection for Reinforcement Learning: Advancing Methods for Evaluation and Detection
by: Nasvytis, Linas, et al.
Published: (2024)

Bayesian Exploration Networks
by: Fellows, Mattie, et al.
Published: (2023)

A Judge-Aware Ranking Framework for Evaluating Large Language Models without Ground Truth
by: Xu, Mingyuan, et al.
Published: (2026)

OpenSanctions Pairs: Large-Scale Entity Matching with LLMs
by: Smith, Chandler, et al.
Published: (2026)

Mirror Learning: A Unifying Framework of Policy Optimisation
by: Kuba, Jakub Grudzien, et al.
Published: (2022)

SAGE: Sparse Adaptive Guidance for Dependency-Aware Tabular Data Generation
by: Yang, Shuo, et al.
Published: (2026)

Towards Principled Evaluations of Sparse Autoencoders for Interpretability and Control
by: Makelov, Aleksandar, et al.
Published: (2024)

Route Sparse Autoencoder to Interpret Large Language Models
by: Shi, Wei, et al.
Published: (2025)

Self-Compatibility: Evaluating Causal Discovery without Ground Truth
by: Faller, Philipp M., et al.
Published: (2023)

Evaluation of Missing Data Imputation for Time Series Without Ground Truth
by: Farjallah, Rania, et al.
Published: (2025)

Sparse Autoencoders Enable Scalable and Reliable Circuit Identification in Language Models
by: O'Neill, Charles, et al.
Published: (2024)

MALT: Improving Reasoning with Multi-Agent LLM Training
by: Motwani, Sumeet Ramesh, et al.
Published: (2024)

Empirical Evaluation of Progressive Coding for Sparse Autoencoders
by: Peter, Hans, et al.
Published: (2025)

Ensembling Sparse Autoencoders
by: Gadgil, Soham, et al.
Published: (2025)

Evaluating Model Explanations without Ground Truth
by: Rawal, Kaivalya, et al.
Published: (2025)

Evaluating Sparse Autoencoders: From Shallow Design to Matching Pursuit
by: Costa, Valérie, et al.
Published: (2025)