Related papers: Model-Agnostic Gender Bias Control for Text-to-Ima…

Neural networks struggle with image classification when biases are learned and misleads correlations, affecting their generalization and performance. Previous methods require attribute labels (e.g. background, color) or utilizes Generative…

Text-to-image diffusion models achieve impressive generation quality but inherit and amplify training-data biases, skewing coverage of semantic attributes. Prior work addresses this in two ways. Closed-set approaches mitigate biases in…

As large language models (LLMs) grow in scale and capability, understanding their internal mechanisms becomes increasingly critical. Sparse autoencoders (SAEs) have emerged as a key tool in mechanistic interpretability, enabling the…

Bias in generative Text-to-Image (T2I) models is a known issue, yet systematically analyzing such models' outputs to uncover it remains challenging. We introduce the Visual Bias Explorer (ViBEx) to interactively explore the output space of…

Sparse autoencoders (SAEs) are widely used in mechanistic interpretability research for large language models; however, the state-of-the-art method of using $k$-sparse autoencoders lacks a theoretical grounding for selecting the…

(Bolukbasi et al., 2016) demonstrated that pretrained word embeddings can inherit gender bias from the data they were trained on. We investigate how this bias affects downstream classification tasks, using the case study of occupation…

Several studies have raised awareness about social biases in image generative models, demonstrating their predisposition towards stereotypes and imbalances. This paper contributes to this growing body of research by introducing an…

Text-to-image (T2I) generation aims at producing realistic images corresponding to text descriptions. Generative Adversarial Network (GAN) has proven to be successful in this task. Typical T2I GANs are 2 phase methods that first pretrain an…

Sparse autoencoders (SAEs) have emerged as powerful techniques for interpretability of large language models (LLMs), aiming to decompose hidden states into meaningful semantic features. While several SAE variants have been proposed, there…

The transformative potential of text-to-image (T2I) models hinges on their ability to synthesize culturally diverse, photorealistic images from textual prompts. However, these models often perpetuate cultural biases embedded within their…

Recent work shows that Sparse Autoencoders (SAE) applied to large language model (LLM) layers have neurons corresponding to interpretable concepts. These SAE neurons can be modified to align generated outputs, but only towards…

Recent advancements in Text-to-Image (T2I) diffusion models have demonstrated impressive success in generating high-quality images with zero-shot generalization capabilities. Yet, current models struggle to closely adhere to prompt…

Dense embeddings deliver strong retrieval performance but often lack interpretability and controllability. This paper introduces a novel approach using sparse autoencoders (SAE) to interpret and control dense embeddings via the learned…

Societal biases present in pre-trained large language models are a critical issue as these models have been shown to propagate biases in countless downstream applications, rendering them unfair towards specific groups of people. Since…

As machine learning-enabled Text-to-Image (TTI) systems are becoming increasingly prevalent and seeing growing adoption as commercial services, characterizing the social biases they exhibit is a necessary first step to lowering their risk…

Text-to-Image (TTI) generative models have shown great progress in the past few years in terms of their ability to generate complex and high-quality imagery. At the same time, these models have been shown to suffer from harmful biases,…

EEG foundation models achieve state-of-the-art clinical performance, yet the internal computations driving their predictions remain opaque: a barrier to clinical trust. We apply TopK Sparse Autoencoders (SAEs) across three architecturally…

Unlearning specific concepts in text-to-image diffusion models has become increasingly important for preventing undesirable content generation. Among prior approaches, sparse autoencoder (SAE)-based methods have attracted attention due to…

Recent advances in text-to-image (T2I) diffusion models have significantly improved the quality of generated images. However, providing efficient control over individual subjects, particularly the attributes characterizing them, remains a…

Sparse Autoencoders (SAEs) have emerged as a promising approach for interpreting neural network representations by learning sparse, human-interpretable features from dense activations. We investigate whether incorporating variational…