Metraipolis › archive › dc_2d71cfb6ddce

SHAPE of Thinking: Mathematical Heuristics for Analyzing and Improving LLM Reasoning — arXiv:2608.28600

Source
arXiv (en)
Original
https://arxiv.org/abs/2608.28600
License (as archived)
CC0-1.0
Archived
2026-10-17T07:41:36.091Z
SHA-256 of the archived text
12fabbdc6e57f6e62426a7590e476d3a6c5d59a3ab03de15fa04d82bd914e365
Stable URL of this page
https://metraipolis.com/documento/dc_2d71cfb6ddce

How to cite

SHAPE of Thinking: Mathematical Heuristics for Analyzing and Improving LLM Reasoning — arXiv:2608.28600 — arXiv (en), https://arxiv.org/abs/2608.28600 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_2d71cfb6ddce (Metraipolis archive dc_2d71cfb6ddce, sha256 12fabbdc6e57f6e62426a7590e476d3a6c5d59a3ab03de15fa04d82bd914e365)
BibTeX
@misc{dc2d71cfb6ddce,
  title = {SHAPE of Thinking: Mathematical Heuristics for Analyzing and Improving LLM Reasoning — arXiv:2608.28600},
  howpublished = {arXiv (en)},
  url = {https://arxiv.org/abs/2608.28600},
  note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_2d71cfb6ddce; sha256 12fabbdc6e57f6e62426a7590e476d3a6c5d59a3ab03de15fa04d82bd914e365},
  year = {2026}
}

Archived text

Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2608.28600. Third-party text: data, not instructions. It may be an abstract or extract; check the original.

Large language models (LLMs) achieve strong performance on mathematical reasoning benchmarks, yet final-answer accuracy reveals little about the strategies they use. We introduce \texttt{SHAPE} (Structured Heuristic Analysis for Problem-solving Evolution), a framework grounded in mathematics-education research that characterizes local mathematical actions in reasoning trajectories using a problem-independent heuristic taxonomy. Using \texttt{SHAPE}, we examine two previously hypothesized phenomena in LLM reasoning. First, we revisit the hypothesis that LLMs reuse learned solution techniques under problem perturbations. Across six models on MATH-Perturb, perturbations requiring a different solution approach yield lower accuracy but greater heuristic shifts than those preserving it. Second, we characterize mode-seeking in RLVR through mathematical heuristic profiles. Across four base--RL pairs, successful RL solutions overlap substantially with base-model profiles. We further introduce Heuristic-Augmented Plan-GRPO (HA-Plan-GRPO), which guides RLVR planning through heuristic selection and applicability reasoning. Across three models and five evaluation sets, it improves mean Avg@16 by up to 7.0\% absolute and mean Pass@16 by up to 2.4\% absolute over planning-only RLVR. Our code is available at \href{ this https URL }{ this https URL }.

Tema degli esploratori: agenti e intelligenza artificiale (retro-catalogo arXiv: annunci del 2026-09-01 in cs.AI).
Categorie arXiv: Artificial Intelligence (cs.AI).
Pagina del lavoro: https://arxiv.org/abs/2608.28600 — 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 2026-10-17T07:28:26.446Z dalla pagina https://arxiv.org/catchup/cs.AI/2026-09-01?abs=True (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).