Related papers: Do Sparse Autoencoders Capture Concept Manifolds?

The manifold hypothesis states that high-dimensional data can be modeled as lying on or near a low-dimensional, nonlinear manifold. Variational Autoencoders (VAEs) approximate this manifold by learning mappings from low-dimensional latent…

Sparse autoencoders (SAEs) are one of the main methods to interpret the inner workings of deep neural networks (DNNs), decomposing activations into higher-dimensional features. However, they exhibit critical shortcomings where a large…

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…

Sparse Autoencoders (SAEs) are a prominent tool in mechanistic interpretability (MI) for decomposing neural network activations into interpretable features. However, the aspiration to identify a canonical set of features is challenged by…

Manifold alignment (MA) involves a set of techniques for learning shared representations across domains, yet many traditional MA methods are incapable of performing out-of-sample extension, limiting their real-world applicability. We…

Sparse autoencoders (SAEs) are a popular technique for interpreting language model activations, and there is extensive recent work on improving SAE effectiveness. However, most prior work evaluates progress using unsupervised proxy metrics…

The central question in representation learning is what constitutes a good or meaningful representation. In this work we argue that if we consider data with inherent cluster structures, where clusters can be characterized through different…

Variational Autoencoders (VAEs) are powerful generative models for learning latent representations. Standard VAEs generate dispersed and unstructured latent spaces by utilizing all dimensions, which limits their interpretability, especially…

Sparse autoencoders (SAEs) \citep{bricken2023monosemanticity,gao2024scalingevaluatingsparseautoencoders} rely on dictionary learning to extract interpretable features from neural networks at scale in an unsupervised manner, with…

Variational Autoencoders are one of the most commonly used generative models, particularly for image data. A prominent difficulty in training VAEs is data that is supported on a lower-dimensional manifold. Recent work by Dai and Wipf (2020)…

We study how reliably sparse autoencoders (SAEs) support claims about reasoning-related internal features in large language models. We first give a stylized analysis showing that sparsity-regularized decoding can preferentially retain…

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…

Syntactic information contains structures and rules about how text sentences are arranged. Incorporating syntax into text modeling methods can potentially benefit both representation learning and generation. Variational autoencoders (VAEs)…

Variational Auto-Encoder (VAE) has been widely applied as a fundamental generative model in machine learning. For complex samples like imagery objects or scenes, however, VAE suffers from the dimensional dilemma between reconstruction…

Linear probes and sparse autoencoders consistently recover meaningful structure from transformer representations -- yet why should such simple methods succeed in deep, nonlinear systems? We show this is not merely an empirical regularity…

Sparse autoencoders (SAEs) are used to disentangle the dense, polysemantic internal representations of large language models (LLMs) into interpretable, monosemantic concepts. However, standard $\ell_1$-regularized SAEs suffer from feature…

The mechanisms behind multilingual capabilities in Large Language Models (LLMs) have been examined using neuron-based or internal-activation-based methods. However, these methods often face challenges such as superposition and layer-wise…

Artistic style transfer in generative models remains a significant challenge, as existing methods often introduce style only via model fine-tuning, additional adapters, or prompt engineering, all of which can be computationally expensive…

LLMs increasingly require surgical model editing to enhance domain-specific capabilities without incurring the computational cost or catastrophic forgetting associated with full fine-tuning. Sparse Autoencoders (SAEs) have emerged as a…

We introduce Sparse Concept Anchoring, a method that biases latent space to position a targeted subset of concepts while allowing others to self-organize, using only minimal supervision (labels for <0.1% of examples per anchored concept).…