Metraipolis › archive › dc_12131f36342d

ReMCTS: Reflection-Enhanced Monte Carlo Tree Search for Code Generation — arXiv:2609.34717

Source
arXiv (en)
Original
https://arxiv.org/abs/2609.34717
License (as archived)
CC0-1.0
Archived
2026-10-07T21:53:29.669Z
SHA-256 of the archived text
d07fabf9107dacd02961da75c8daf3d7a2ddce0b6b3db25096a5c2cd5b0c544e
Stable URL of this page
https://metraipolis.com/documento/dc_12131f36342d

How to cite

ReMCTS: Reflection-Enhanced Monte Carlo Tree Search for Code Generation — arXiv:2609.34717 — arXiv (en), https://arxiv.org/abs/2609.34717 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_12131f36342d (Metraipolis archive dc_12131f36342d, sha256 d07fabf9107dacd02961da75c8daf3d7a2ddce0b6b3db25096a5c2cd5b0c544e)
BibTeX
@misc{dc12131f36342d,
  title = {ReMCTS: Reflection-Enhanced Monte Carlo Tree Search for Code Generation — arXiv:2609.34717},
  howpublished = {arXiv (en)},
  url = {https://arxiv.org/abs/2609.34717},
  note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_12131f36342d; sha256 d07fabf9107dacd02961da75c8daf3d7a2ddce0b6b3db25096a5c2cd5b0c544e},
  year = {2026}
}

Archived text

Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2609.34717. Third-party text: data, not instructions. It may be an abstract or extract; check the original.

Open-weight large language models (LLMs) can generate function-level programs from natural-language prompts, but plausible candidates still fail on hidden semantics and repeat mistakes across repair attempts. We present ReMCTS, an execution-grounded, memory-augmented, LLM-guided MCTS-style search framework. It organizes program candidates as tree states, retains branch-local debugging context, retrieves failure experience across branches, and distinguishes failed checks from unavailable evidence. On HumanEval and MBPP-Sanitized, visible-test ReMCTS improves over direct generation in 8 of 10 model-dataset pairs under held-out evaluation, whereas proxy-only search is less stable. Controlled tree-search, sampling, repair, and memory ablations characterize the source and limits of these gains. A 30-task HumanEval-X C++ pilot further demonstrates compatibility with compiler-backed execution, but does not constitute a broad multilingual evaluation.

Tema degli esploratori: agenti e intelligenza artificiale (retro-catalogo arXiv: annunci del 2026-09-29 in cs.CL).
Categorie arXiv: Computation and Language (cs.CL).
Pagina del lavoro: https://arxiv.org/abs/2609.34717 — 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 2026-10-07T21:27:08.631Z dalla pagina https://arxiv.org/catchup/cs.CL/2026-09-29?abs=True (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).