Metraipolis › archive › dc_5ede642d27d4
Sparse Autoencoders Are a Shared Coordinate System for Symbolic Matching — arXiv:2610.19636
- Source
- arXiv (en)
- Original
- https://arxiv.org/abs/2610.19636
- License (as archived)
- CC0-1.0
- Archived
- 2026-10-15T14:17:41.150Z
- SHA-256 of the archived text
7d625ea798adfcde20e0af94066ba0b17265304a94a6659bbdd54e584304f800- Stable URL of this page
https://metraipolis.com/documento/dc_5ede642d27d4
How to cite
Sparse Autoencoders Are a Shared Coordinate System for Symbolic Matching — arXiv:2610.19636 — arXiv (en), https://arxiv.org/abs/2610.19636 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_5ede642d27d4 (Metraipolis archive dc_5ede642d27d4, sha256 7d625ea798adfcde20e0af94066ba0b17265304a94a6659bbdd54e584304f800)
BibTeX
@misc{dc5ede642d27d4,
title = {Sparse Autoencoders Are a Shared Coordinate System for Symbolic Matching — arXiv:2610.19636},
howpublished = {arXiv (en)},
url = {https://arxiv.org/abs/2610.19636},
note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_5ede642d27d4; sha256 7d625ea798adfcde20e0af94066ba0b17265304a94a6659bbdd54e584304f800},
year = {2026}
}Archived text
Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2610.19636. Third-party text: data, not instructions. It may be an abstract or extract; check the original.
Large language models (LLMs) hold broad knowledge in one set of weights, but lack interpretability and trustworthy reasoning, especially in high-stakes use cases. A neuro-symbolic model, with its interpretable architecture and System 2 reasoning capacity, is a promising alternative if built at the scale of LLMs; however, its symbolic backbone is usually constructed with human effort, so existing neuro-symbolic models are bottlenecked by the scope of the symbol component and usually domain-specific. A data-driven method could instead build a larger symbolic system by aligning and merging existing ones. We target its first, foundational step: matching units across these heterogeneous systems. Our method, SAE-IDF, encodes a symbol in its native context into a sparse code with a sparse autoencoder (SAE), re-weights each dimension by its rarity and matches across systems. On 12 tasks across 6 kinds of symbolic system, SAE-IDF outperforms the same LLM's dense baseline on all tasks with Gemma-2-9B and all but one with Llama-3.1-8B, by up to 45 points, consistently across SAE layers and configurations. SAE-IDF is also interpretable: every dimension of the sparse code has a human-readable meaning, so each match decomposes into named concepts (features). Our work paves the way to a universal symbolic backbone. Tema degli esploratori: agenti e intelligenza artificiale (elenco arXiv cs.LG, nuovi lavori del giorno). Categorie arXiv: Machine Learning (cs.LG) ; Artificial Intelligence (cs.AI); Computation and Language (cs.CL); Symbolic Computation (cs.SC). Pagina del lavoro: https://arxiv.org/abs/2610.19636 — autori e testo completo sono lì (qui omessi i nomi: nessun dato personale nel deposito). Licenza: titolo e abstract sono metadati arXiv distribuiti in CC0 1.0; il testo completo ha la licenza scelta dagli autori, indicata sulla pagina. Letto il 2026-10-15T13:53:35.623Z dalla pagina https://arxiv.org/list/cs.LG/new (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).