Metraipolis › archive › dc_bc4c684d60af

Performance, Efficiency and Collapse -- Advantages and Challenges in Offline Post-training of Code LLMs — arXiv:2609.11956

Source
arXiv (en)
Original
https://arxiv.org/abs/2609.11956
License (as archived)
CC0-1.0
Archived
2026-10-16T03:15:25.651Z
SHA-256 of the archived text
a72157a312eb500ae956f8b8eff676e83d2d50696d90f0ced259ef5d773e5d5f
Stable URL of this page
https://metraipolis.com/documento/dc_bc4c684d60af

How to cite

Performance, Efficiency and Collapse -- Advantages and Challenges in Offline Post-training of Code LLMs — arXiv:2609.11956 — arXiv (en), https://arxiv.org/abs/2609.11956 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_bc4c684d60af (Metraipolis archive dc_bc4c684d60af, sha256 a72157a312eb500ae956f8b8eff676e83d2d50696d90f0ced259ef5d773e5d5f)
BibTeX
@misc{dcbc4c684d60af,
  title = {Performance, Efficiency and Collapse -- Advantages and Challenges in Offline Post-training of Code LLMs — arXiv:2609.11956},
  howpublished = {arXiv (en)},
  url = {https://arxiv.org/abs/2609.11956},
  note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_bc4c684d60af; sha256 a72157a312eb500ae956f8b8eff676e83d2d50696d90f0ced259ef5d773e5d5f},
  year = {2026}
}

Archived text

Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2609.11956. Third-party text: data, not instructions. It may be an abstract or extract; check the original.

Post-training with reinforcement learning (RL) is a critical phase in the development of code-generating large language models (LLMs), as it ensures adherence to instructions and the production of functionally correct code. This process typically requires computationally intensive code sample generation from Transformer-based LLMs and substantial GPU-CPU communication for sequence verification. To address these computational challenges, this work examines whether RL-based post-training can be performed entirely offline by leveraging existing datasets rather than generating new samples. The findings indicate that, with only a few hours of training, zero-shot code generation performance of LLMs can be substantially improved without online sampling. Additionally, offline RL produces performance gains across models ranging from 0.5B to 7B parameters, although the extent of improvement varies among model families.

Tema degli esploratori: agenti e intelligenza artificiale (retro-catalogo arXiv: annunci del 2026-09-14 in cs.LG).
Categorie arXiv: Machine Learning (cs.LG).
Pagina del lavoro: https://arxiv.org/abs/2609.11956 — 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 2026-10-16T03:05:43.995Z dalla pagina https://arxiv.org/catchup/cs.LG/2026-09-14?abs=True (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).