SMIXAE: Towards Unsupervised Manifold Discovery in Language Models

• URL: http://arxiv.org/abs/2605.09224v1
• Date: Sat, 09 May 2026 23:48:27 GMT
• Title: SMIXAE: Towards Unsupervised Manifold Discovery in Language Models
• Abstract summary: Sparse autoencoders (SAEs) have been used widely to decompose and interpret neural network activations.<n>One key issue with SAEs is their inability to directly model multidimensional features.<n>We begin to address this issue by introducing the Sparse MIXture of Autoencoders architecture.
• Score: 0.0
• License: http://creativecommons.org/licenses/by-nc-sa/4.0/
• Abstract: Sparse autoencoders (SAEs) have been used widely to decompose and interpret neural network activations, especially those of transformer language models. One key issue with SAEs is their inability to directly model multidimensional features. Instead, SAEs may tile such features by a set of independent directions that must be grouped together after the SAE training phase, impeding discoverability and interpretation of learned feature representations. We begin to address this issue by introducing the Sparse MIXture of Autoencoders (SMIXAE) architecture. Empirically, we provide evidence that SMIXAE models have success both in directly learning previously identified manifold structures, as well as finding novel structures, within the open source Gemma 2 2B and 9B models. Finally, we discuss several limitations and point towards areas for future work.

Related papers

• Mechanistic Interpretability of ASR models using Sparse Autoencoders [2.835219195516112]
  Internal mechanisms of deep Transformer-based NLP models remain largely a mystery.<n>Sparse Autoencoders (SAE) have emerged to understand these mechanisms by projecting dense representations into a sparse vector.<n>In this work, a SAE is applied to Whisper, a Transformer-based ASR, training a high-dimensional sparse latent space on frame-level embeddings extracted from the Whisper encoder.
  arXiv  Detail & Related papers  (2026-05-12T15:02:20Z)
• Qwen-Scope: Turning Sparse Features into Development Tools for Large Language Models [80.45129499188461]
  We introduce Qwen-Scope, an open-source suite of sparse autoencoders (SAEs) built on the Qwen model family.<n>We show that SAEs can go beyond post-hoc analysis to serve as practical interfaces for model development along four directions.
  arXiv  Detail & Related papers  (2026-05-12T10:01:06Z)
• Beyond Redundancy: Diverse and Specialized Multi-Expert Sparse Autoencoder [59.89996751196727]
  Sparse autoencoders (SAEs) have emerged as a powerful tool for interpreting large language models.<n>SAEs' hidden layers have high dimensionality to satisfy sparsity constraints, resulting in prohibitive training and inference costs.<n>Recent Mixture of Experts (MoE) approaches attempt to address this by SAEs into narrower expert networks with gated activation.<n>We propose two key innovations: (1) Multiple Expert Activation that simultaneously engages semantically weighted expert subsets to encourage specialization, and (2) Feature Scaling that enhances diversity through adaptive high-frequency scaling.
  arXiv  Detail & Related papers  (2025-11-07T22:19:34Z)
• Ensembling Sparse Autoencoders [10.81463830315253]
  Sparse autoencoders (SAEs) are used to decompose neural network activations into human-interpretable features.<n>We propose to ensemble multiple SAEs through naive bagging and boosting.<n>Our empirical results demonstrate that ensembling SAEs can improve the reconstruction of language model activations, diversity of features, and SAE stability.
  arXiv  Detail & Related papers  (2025-05-21T23:31:21Z)
• Stack Transformer Based Spatial-Temporal Attention Model for Dynamic Sign Language and Fingerspelling Recognition [1.949837893170278]
  Hand gesture-based Sign Language Recognition serves as a crucial bridge between deaf and non-deaf individuals.<n>We propose the Sequential Spatio-Temporal Attention Network (SSTAN), a novel Transformer-based architecture.<n>We validated our model through extensive experiments on diverse, large-scale datasets.
  arXiv  Detail & Related papers  (2025-03-21T04:57:18Z)
• Interpreting CLIP with Hierarchical Sparse Autoencoders [8.692675181549117]
  Matryoshka SAE (MSAE) learns hierarchical representations at multiple granularities simultaneously.<n>MSAE establishes a new state-of-the-art frontier between reconstruction quality and sparsity for CLIP.
  arXiv  Detail & Related papers  (2025-02-27T22:39:13Z)
• Transformers Use Causal World Models in Maze-Solving Tasks [49.67445252528868]
  We identify World Models in transformers trained on maze-solving tasks.<n>We find that it is easier to activate features than to suppress them.<n> positional encoding schemes appear to influence how World Models are structured within the model's residual stream.
  arXiv  Detail & Related papers  (2024-12-16T15:21:04Z)
• One-Step is Enough: Sparse Autoencoders for Text-to-Image Diffusion Models [31.493126840872417]
  We train SAEs to learn interpretable features for SDXL Turbo, a few-step text-to-image diffusion model.<n>We show that their learned features are interpretable, causally influence the generation process, and reveal specialization among the blocks.<n>Our work is the first investigation of SAEs for interpretability in text-to-image diffusion models.
  arXiv  Detail & Related papers  (2024-10-28T19:01:18Z)
• Llama Scope: Extracting Millions of Features from Llama-3.1-8B with Sparse Autoencoders [115.34050914216665]
  Sparse Autoencoders (SAEs) have emerged as a powerful unsupervised method for extracting sparse representations from language models. We introduce a suite of 256 SAEs, trained on each layer and sublayer of the Llama-3.1-8B-Base model, with 32K and 128K features. We assess the generalizability of SAEs trained on base models to longer contexts and fine-tuned models.
  arXiv  Detail & Related papers  (2024-10-27T17:33:49Z)
• Self-Supervised Open-Ended Classification with Small Visual Language Models [60.23212389067007]
  We present Self-Context Adaptation (SeCAt), a self-supervised approach that unlocks few-shot abilities for open-ended classification with small visual language models. By using models with approximately 1B parameters we outperform the few-shot abilities of much larger models, such as Frozen and FROMAGe.
  arXiv  Detail & Related papers  (2023-09-30T21:41:21Z)
• Decoupled Multi-task Learning with Cyclical Self-Regulation for Face Parsing [71.19528222206088]
  We propose a novel Decoupled Multi-task Learning with Cyclical Self-Regulation for face parsing. Specifically, DML-CSR designs a multi-task model which comprises face parsing, binary edge, and category edge detection. Our method achieves the new state-of-the-art performance on the Helen, CelebA-HQ, and LapaMask datasets.
  arXiv  Detail & Related papers  (2022-03-28T02:12:30Z)
• Sign Language Recognition via Skeleton-Aware Multi-Model Ensemble [71.97020373520922]
  Sign language is commonly used by deaf or mute people to communicate. We propose a novel Multi-modal Framework with a Global Ensemble Model (GEM) for isolated Sign Language Recognition ( SLR) Our proposed SAM- SLR-v2 framework is exceedingly effective and achieves state-of-the-art performance with significant margins.
  arXiv  Detail & Related papers  (2021-10-12T16:57:18Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.