cs.LGNov 16, 2026

Representational Change Under Self-Consuming Training: A Sparse Autoencoder Analysis

Authors: Suchit Gupte, Zhongteng Cai, Xueru Zhang, Mohammad Mahdi Khalili

Organizations: Department of Computer Science, The Ohio State University, Columbus, OH 43210, USA

Abstract

Large language models are increasingly exposed to model-generated text, raising the possibility that future models will be trained recursively on the outputs of their predecessors. Prior work on such self-consuming training loops has largely focused on output quality and downstream performance, leaving their effects on internal representations poorly understood. We study how representations evolve across five generations of self-consumption and whether sparse autoencoders (SAEs) remain reliable under this shift. At each generation, we compare a lineage trained on synthetic data with a matched control trained on real data under the same training settings. We evaluate how well SAEs trained before self-consumption transfer to later generations, measuring reconstruction and sparsity across SAE widths, architectures, and model layers. We then fine-tune the original SAE on each generation and use sparse probing to test whether task-relevant information remains concentrated in a small number of latents or becomes more distributed across the representation. We find that zero-shot SAE performance degrades progressively under self-consumption, with growing reconstruction and sparsity gaps that vary substantially across layers. Fine-tuning largely restores reconstruction, but not feature-level usefulness: task-relevant information remains accessible in the full latent space while becoming less concentrated in individual features. These results show that self-consumption induces representational drift and that reconstruction fidelity alone is insufficient to establish that an adapted SAE remains equally useful for interpretation.

Figures & tables

Appendix figures & tables2 assets

Supplementary material from the paper’s appendix.

Appendix

Explore similar work

CardsList
  1. Sparse but Wrong: Incorrect L0 Leads to Incorrect Features in Sparse Autoencoders

    Aug 22, 2025David Chanin, Adrià Garriga-AlonsoActivation SparsityLLM Interpretability

  2. Diminishing Returns of Data in Sparse Autoencoders: Functional Behavior does not Guarantee Feature Reproducibility

    Oct 14, 2026Sambit Ghosh, Madelyn Mathai, R. Venkatesh Babu +1Sparse AutoencodersData-Efficient Learning

  3. Representation Redistribution in Sparse Autoencoders

    Oct 15, 2026Ziye Mei, Haotian MaSparse AutoencodersPrompt Sensitivity