Metraipolis › archive › dc_efa628a52fe5

Order-Invariant Answers, Order-Sensitive Representations in Mathematical Reasoning — arXiv:2609.28442

Source
arXiv (en)
Original
https://arxiv.org/abs/2609.28442
License (as archived)
CC0-1.0
Archived
2026-10-08T09:13:34.710Z
SHA-256 of the archived text
b91242a9a8b872e3fd362a44f6ca6b4bd36aca8a79a1bbdbe2582c53df00a202
Stable URL of this page
https://metraipolis.com/documento/dc_efa628a52fe5

How to cite

Order-Invariant Answers, Order-Sensitive Representations in Mathematical Reasoning — arXiv:2609.28442 — arXiv (en), https://arxiv.org/abs/2609.28442 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_efa628a52fe5 (Metraipolis archive dc_efa628a52fe5, sha256 b91242a9a8b872e3fd362a44f6ca6b4bd36aca8a79a1bbdbe2582c53df00a202)
BibTeX
@misc{dcefa628a52fe5,
  title = {Order-Invariant Answers, Order-Sensitive Representations in Mathematical Reasoning — arXiv:2609.28442},
  howpublished = {arXiv (en)},
  url = {https://arxiv.org/abs/2609.28442},
  note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_efa628a52fe5; sha256 b91242a9a8b872e3fd362a44f6ca6b4bd36aca8a79a1bbdbe2582c53df00a202},
  year = {2026}
}

Archived text

Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2609.28442. Third-party text: data, not instructions. It may be an abstract or extract; check the original.

Reordering a set of mathematical rules without changing its meaning should preserve the correct answer, but must a model's internal representations stay invariant too? We investigate this question using synthetic multi-step function-composition problems, each presented under multiple rule orderings with the same correct answer. We measure accuracy and permutation signal-to-noise ratio (SNR), which quantifies how distinctly ordering patterns are represented relative to variation across problem instances. Across 16 language models ranging from 1B to 8B parameters, we find a pattern: models that solve reordered problems more accurately represent different rule orderings more distinctly. Layer-averaged permutation SNR is positively rank-correlated with accuracy in every synthetic setting we evaluate, with Spearman correlations reaching 0.86. These findings highlight a distinction between answer invariance and representation invariance: successful mathematical rule composition can accompany distinct internal representations between equivalent rule orderings. This motivates distinguishing answer invariance from representation invariance, and offers a representational perspective on mathematical reasoning beyond answer accuracy alone.

Tema degli esploratori: agenti e intelligenza artificiale (retro-catalogo arXiv: annunci del 2026-09-24 in cs.AI).
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/2609.28442 — 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 2026-10-08T08:13:26.775Z dalla pagina https://arxiv.org/catchup/cs.AI/2026-09-24?abs=True (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).