English
Related papers

Related papers: Sparse Autoencoders Make Audio Foundation Models m…

200 papers

Inspite the emerging importance of Speech Emotion Recognition (SER), the state-of-the-art accuracy is quite low and needs improvement to make commercial applications of SER viable. A key underlying reason for the low accuracy is the…

Sound · Computer Science 2020-03-24 Siddique Latif , Rajib Rana , Sara Khalifa , Raja Jurdak , Julien Epps , Björn W. Schuller

Self-supervised learning has been a powerful training paradigm to facilitate representation learning. In this study, we design a masked autoencoder (MAE) to guide deep learning models to learn electroencephalography (EEG) signal…

Human-Computer Interaction · Computer Science 2024-09-04 Yifei Zhou , Sitong Liu

Sparse autoencoders (SAEs) are widely used in mechanistic interpretability to project LLM activations onto sparse latent spaces. However, sparsity alone is an imperfect proxy for interpretability, and current training objectives often…

Machine Learning · Computer Science 2026-04-09 Vivek Narayanaswamy , Kowshik Thopalli , Bhavya Kailkhura , Wesam Sakla

Bioacoustic recognition requires fine-grained acoustic understanding to distinguish similar-sounding species. However, many large-scale data repositories such as iNaturalist are weakly annotated, often with only a single positive species…

Sound · Computer Science 2026-05-15 Wuao Liu , Mustafa Chasmai , Subhransu Maji , Grant Van Horn

In this paper, we propose to pre-train audio encoders using synthetic patterns instead of real audio data. Our proposed framework consists of two key elements. The first one is Masked Autoencoder (MAE), a self-supervised learning framework…

Audio and Speech Processing · Electrical Eng. & Systems 2024-10-02 Yuchi Ishikawa , Tatsuya Komatsu , Yoshimitsu Aoki

Responsible deployment of language models requires mechanisms for refusing unsafe prompts while preserving model performance. While most approaches modify model weights through additional training, we explore an alternative: steering model…

In speech recognition, it is essential to model the phonetic content of the input signal while discarding irrelevant factors such as speaker variations and noise, which is challenging in low-resource settings. Self-supervised pre-training…

Computation and Language · Computer Science 2023-01-04 Sreepratha Ram , Hanan Aldarmaki

Recently, a variational autoencoder (VAE)-based single-channel speech enhancement system using Bayesian permutation training has been proposed, which uses two pretrained VAEs to obtain latent representations for speech and noise. Based on…

Audio and Speech Processing · Electrical Eng. & Systems 2026-02-03 Jiatong Li , Simon Doclo

Automatic singing voice understanding tasks, such as singer identification, singing voice transcription, and singing technique classification, benefit from data-driven approaches that utilize deep learning techniques. These approaches work…

Sound · Computer Science 2023-09-06 Yuya Yamamoto

Sparse autoencoders (SAEs) are a promising unsupervised approach for identifying causally relevant and interpretable linear features in a language model's (LM) activations. To be useful for downstream tasks, SAEs need to decompose LM…

Machine Learning · Computer Science 2024-08-02 Senthooran Rajamanoharan , Tom Lieberum , Nicolas Sonnerat , Arthur Conmy , Vikrant Varma , János Kramár , Neel Nanda

An ability to model a generative process and learn a latent representation for speech in an unsupervised fashion will be crucial to process vast quantities of unlabelled speech data. Recently, deep probabilistic generative models such as…

Computation and Language · Computer Science 2017-09-25 Wei-Ning Hsu , Yu Zhang , James Glass

Self-supervised representation learning approaches have grown in popularity due to the ability to train models on large amounts of unlabeled data and have demonstrated success in diverse fields such as natural language processing, computer…

Machine Learning · Computer Science 2023-02-06 John Harvill , Jarred Barber , Arun Nair , Ramin Pishehvar

Recent general-purpose audio representations show state-of-the-art performance on various audio tasks. These representations are pre-trained by self-supervised learning methods that create training signals from the input. For example,…

Audio and Speech Processing · Electrical Eng. & Systems 2023-03-09 Daisuke Niizumi , Daiki Takeuchi , Yasunori Ohishi , Noboru Harada , Kunio Kashino

Sparse Autoencoder (SAE) features have become essential tools for mechanistic interpretability research. SAE features are typically characterized by examining their activating examples, which are often "monosemantic" and align with human…

Artificial Intelligence · Computer Science 2025-09-30 Claire Tian , Katherine Tian , Nathan Hu

Sparse autoencoders (SAEs) model the activations of a neural network as linear combinations of sparsely occurring directions of variation (latents). The ability of SAEs to reconstruct activations follows scaling laws w.r.t. the number of…

Machine Learning · Computer Science 2025-09-05 Eric J. Michaud , Liv Gorton , Tom McGrath

Large Language Models (LLMs) have achieved strong complex reasoning capabilities through Chain-of-Thought (CoT) reasoning. However, their reasoning patterns remain too complicated to analyze. While Sparse Autoencoders (SAEs) have emerged as…

Machine Learning · Computer Science 2026-03-04 Xuan Yang , Jiayu Liu , Yuhang Lai , Hao Xu , Zhenya Huang , Ning Miao

Sparse autoencoders (SAEs) have proven effective for extracting monosemantic features from large language models (LLMs), yet these features are typically identified in isolation. However, broad evidence suggests that LLMs capture the…

Artificial Intelligence · Computer Science 2026-02-13 Yifan Luo , Yang Zhan , Jiedong Jiang , Tianyang Liu , Mingrui Wu , Zhennan Zhou , Bin Dong

In this work, we demonstrate that affine mappings between residual streams of language models is a cheap way to effectively transfer represented features between models. We apply this technique to transfer the weights of Sparse Autoencoders…

Computation and Language · Computer Science 2025-11-04 Alan Chen , Jack Merullo , Alessandro Stolfo , Ellie Pavlick

Automatic speaker verification (ASV) systems are highly vulnerable to presentation attacks, also called spoofing attacks. Replay is among the simplest attacks to mount - yet difficult to detect reliably. The generalization failure of…

Audio and Speech Processing · Electrical Eng. & Systems 2020-03-24 Bhusan Chettri , Tomi Kinnunen , Emmanouil Benetos

This paper introduces an efficient and robust method for discovering interpretable circuits in large language models using discrete sparse autoencoders. Our approach addresses key limitations of existing techniques, namely computational…

Computation and Language · Computer Science 2024-05-22 Charles O'Neill , Thang Bui