Cover Image

Probing the Representational Power of Sparse Autoencoders in Vision Models

Fuente: arXiv

Saved in:

Bibliographic Details
Main Authors:	Olson, Matthew Lyle, Hinck, Musashi, Ratzlaff, Neale, Li, Changbai, Howard, Phillip, Lal, Vasudev, Tseng, Shao-Yen
Format:	Preprint
Published:	2025
Subjects:	Computer Vision and Pattern Recognition Machine Learning
Online Access:	Acceder al recurso
Tags:	Add Tag No Tags, Be the first to tag this record!

Similar Items

Analyzing Hierarchical Structure in Vision Models with Sparse Autoencoders
by: Olson, Matthew Lyle, et al.
Published: (2025)

Debiasing Large Vision-Language Models by Ablating Protected Attribute Representations
by: Ratzlaff, Neale, et al.
Published: (2024)

Debias your Large Multi-Modal Model at Test-Time via Non-Contrastive Visual Attribute Steering
by: Ratzlaff, Neale, et al.
Published: (2024)

Steering Large Language Models to Evaluate and Amplify Creativity
by: Olson, Matthew Lyle, et al.
Published: (2024)

Training-Free Mitigation of Language Reasoning Degradation After Multimodal Instruction Tuning
by: Ratzlaff, Neale, et al.
Published: (2024)

Why do LLaVA Vision-Language Models Reply to Images in English?
by: Hinck, Musashi, et al.
Published: (2024)

Probing Semantic Routing in Large Mixture-of-Expert Models
by: Olson, Matthew Lyle, et al.
Published: (2025)

LieCraft: A Multi-Agent Framework for Evaluating Deceptive Capabilities in Language Models
by: Olson, Matthew Lyle, et al.
Published: (2026)

LLaVA-Gemma: Accelerating Multimodal Foundation Models with a Compact Language Model
by: Hinck, Musashi, et al.
Published: (2024)

Cultural Awareness in Vision-Language Models: A Cross-Country Exploration
by: Madasu, Avinash, et al.
Published: (2025)

LVLM-Interpret: An Interpretability Tool for Large Vision-Language Models
by: Stan, Gabriela Ben Melech, et al.
Published: (2024)

ClimDetect: A Benchmark Dataset for Climate Change Detection and Attribution
by: Yu, Sungduk, et al.
Published: (2024)

DPO Learning with LLMs-Judge Signal for Computer Use Agents
by: Luo, Man, et al.
Published: (2025)

Pruning the Paradox: How CLIP's Most Informative Heads Enhance Performance While Amplifying Bias
by: Madasu, Avinash, et al.
Published: (2025)

SocialCounterfactuals: Probing and Mitigating Intersectional Social Biases in Vision-Language Models with Counterfactual Examples
by: Howard, Phillip, et al.
Published: (2023)

Quantifying and Enabling the Interpretability of CLIP-like Models
by: Madasu, Avinash, et al.
Published: (2024)

FiVL: A Framework for Improved Vision-Language Alignment through the Lens of Training, Evaluation and Explainability
by: Aflalo, Estelle, et al.
Published: (2024)

L-MAGIC: Language Model Assisted Generation of Images with Coherence
by: Cai, Zhipeng, et al.
Published: (2024)

ICSVR: Investigating Compositional and Syntactic Understanding in Video Retrieval Models
by: Madasu, Avinash, et al.
Published: (2023)

SK-VQA: Synthetic Knowledge Generation at Scale for Training Context-Augmented Multimodal LLMs
by: Su, Xin, et al.
Published: (2024)

LVLM-Compress-Bench: Benchmarking the Broader Impact of Large Vision-Language Model Compression
by: Kundu, Souvik, et al.
Published: (2025)

BridgeTower: Building Bridges Between Encoders in Vision-Language Representation Learning
by: Xu, Xiao, et al.
Published: (2022)

Cultural Counterfactuals: Evaluating Cultural Biases in Large Vision-Language Models with Counterfactual Examples
by: Howard, Phillip, et al.
Published: (2026)

Interpretable and Testable Vision Features via Sparse Autoencoders
by: Stevens, Samuel, et al.
Published: (2025)

Uncovering Bias in Large Vision-Language Models at Scale with Counterfactuals
by: Howard, Phillip, et al.
Published: (2024)

Interpretability Transfer from Language to Vision via Sparse Autoencoders
by: Kravets, Alexey, et al.
Published: (2026)

Causal Interpretation of Sparse Autoencoder Features in Vision
by: Han, Sangyu, et al.
Published: (2025)

SAUCE: Selective Concept Unlearning in Vision-Language Models with Sparse Autoencoders
by: Li, Qing, et al.
Published: (2025)

Sparse Autoencoders for Interpretable Medical Image Representation Learning
by: Wesp, Philipp, et al.
Published: (2026)

Sparse Autoencoders are Topic Models
by: Girrbach, Leander, et al.
Published: (2025)

Universal Sparse Autoencoders: Interpretable Cross-Model Concept Alignment
by: Thasarathan, Harrish, et al.
Published: (2025)

Uncovering Bias in Large Vision-Language Models with Counterfactuals
by: Howard, Phillip, et al.
Published: (2024)

Sparse Autoencoders Learn Monosemantic Features in Vision-Language Models
by: Pach, Mateusz, et al.
Published: (2025)

Cross-Cultural Value Awareness in Large Vision-Language Models
by: Howard, Phillip, et al.
Published: (2026)

Surf3R: Rapid Surface Reconstruction from Sparse RGB Views in Seconds
by: Zhu, Haodong, et al.
Published: (2025)

Why Far Looks Up: Probing Spatial Representation in Vision-Language Models
by: Min, Cheolhong, et al.
Published: (2026)

SparseOcc: Rethinking Sparse Latent Representation for Vision-Based Semantic Occupancy Prediction
by: Tang, Pin, et al.
Published: (2024)

Investigating Zero-Shot Diagnostic Pathology in Vision-Language Models with Efficient Prompt Design
by: Sharma, Vasudev, et al.
Published: (2025)

Improving Reconstruction of Representation Autoencoder
by: Liu, Siyu, et al.
Published: (2026)

From Diagnosis to Improvement: Probing Spatio-Physical Reasoning in Vision Language Models
by: Han, Tiancheng, et al.
Published: (2025)