paper-with-me

Papers

An X-Ray Is Worth 15 Features: Sparse Autoencoders for Interpretable Radiology Report Generation

2024-10-04 · Ahmed Abdulaal, Hugo Fry, Nina Montaña-Brown, Ayodeji Ijishakin, Jack Gao, Stephanie Hyland, Daniel C. Alexander, Daniel C. Castro

Radiological services are experiencing unprecedented demand, leading to increased interest in automating radiology report generation. Existing Vision-Language Models (VLMs) suffer from hallucinations, lack interpretability, and require expensive fine-tuning. We introduce SAE-Rad, which uses sparse autoencoders (SAEs) to decompose latent representations from a pre-trained vision transformer into human-interpretable features. Our hybrid architecture combines state-of-the-art SAE advancements, achieving accurate latent reconstructions while maintaining sparsity. Using an off-the-shelf language model, we distil ground-truth reports into radiological descriptions for each SAE feature, which we then compile into a full report for each image, eliminating the need for fine-tuning large models for this task. To the best of our knowledge, SAE-Rad represents the first instance of using mechanistic interpretability techniques explicitly for a downstream multi-modal reasoning task. On the MIMIC-CXR dataset, SAE-Rad achieves competitive radiology-specific metrics compared to state-of-the-art models while using significantly fewer computational resources for training. Qualitative analysis reveals that SAE-Rad learns meaningful visual concepts and generates reports aligning closely with expert interpretations. Our results suggest that SAEs can enhance multimodal reasoning in healthcare, providing a more interpretable alternative to existing VLMs.

📄 PDF Abstract BibTeX arXiv:2410.03334

Code (0)

등록된 구현이 없습니다.

Tasks

Language ModellingMultimodal Reasoning

Methods 이 논문이 사용한 방법론

Linear Layer A Linear Layer is a projection $\mathbf{XW + b}$.
Softmax The Softmax output function transforms a previous layer's output into a vector of probabilities. It is commonly used for multiclass classification. Given an input vector $x$…
Multi-Head Attention 설명 없음
Layer Normalization Unlike batch normalization, Layer Normalization directly estimates the normalization statistics from the summed inputs…
Dense Connections Dense Connections, or Fully Connected Connections, are a type of layer in a deep neural network that use a linear operation where every input is connected to every output…
Residual Connection 설명 없음
Attention 설명 없음
Vision Transformer The Vision Transformer, or ViT, is a model for image classification that employs a Transformer-like architecture over…

Similar Papers 제목 키워드 기반

Insights into a radiology-specialised multimodal large language model with sparse autoencoders

2025-07-17 · Kenza Bouzid, Shruthi Bannur, Felix Meissen, Daniel Coelho de Castro 외 arxiv

Interpretability can improve the safety, transparency and trust of AI models, which is especially important in healthcare applications where decisions often carry significant consequences. Mechanistic interpretability, p…

TIDE : Temporal-Aware Sparse Autoencoders for Interpretable Diffusion Transformers in Image Generation

2025-03-10 · Victor Shea-Jay Huang, Le Zhuo, Yi Xin, Zhaokai Wang 외

Diffusion Transformers (DiTs) are a powerful yet underexplored class of generative models compared to U-Net-based diffusion models. To bridge this gap, we introduce TIDE (Temporal-aware Sparse Autoencoders for Interpreta…

DenoisingImage GenerationStyle Transfer

Interpreting CFD Surrogates through Sparse Autoencoders

2025-07-21 · Yeping Hu, Shusen Liu arxiv

Learning-based surrogate models have become a practical alternative to high-fidelity CFD solvers, but their latent representations remain opaque and hinder adoption in safety-critical or regulation-bound settings. This w…

Exploring Task Performance with Interpretable Models via Sparse Auto-Encoders

2025-07-08 · Shun Wang, Tyler Loakman, Youbo Lei, Yi Liu 외 arxiv

Large Language Models (LLMs) are traditionally viewed as black-box algorithms, therefore reducing trustworthiness and obscuring potential approaches to increasing performance on downstream tasks. In this work, we apply a…

Mathematical Reasoning

Efficient Dictionary Learning with Switch Sparse Autoencoders

2024-10-10 · Anish Mudide, Joshua Engels, Eric J. Michaud, Max Tegmark 외

Sparse autoencoders (SAEs) are a recent technique for decomposing neural network activations into human-interpretable features. However, in order for SAEs to identify all features represented in frontier models, it will …

Dictionary LearningMixture-of-Experts