Metraipolis › archive › dc_28b586c2db0a
Ladders in Chaos: When, How, (and Perhaps Why) Does Test-Time Scaling Improve LLM Machine Translation — arXiv:2608.28496
- Source
- arXiv (en)
- Original
- https://arxiv.org/abs/2608.28496
- License (as archived)
- CC0-1.0
- Archived
- 2026-10-17T10:56:39.667Z
- SHA-256 of the archived text
70037357890b5ffb13f888470575ce44c892750e8a53e3cc5c7523668026e999- Stable URL of this page
https://metraipolis.com/documento/dc_28b586c2db0a
How to cite
Ladders in Chaos: When, How, (and Perhaps Why) Does Test-Time Scaling Improve LLM Machine Translation — arXiv:2608.28496 — arXiv (en), https://arxiv.org/abs/2608.28496 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_28b586c2db0a (Metraipolis archive dc_28b586c2db0a, sha256 70037357890b5ffb13f888470575ce44c892750e8a53e3cc5c7523668026e999)
BibTeX
@misc{dc28b586c2db0a,
title = {Ladders in Chaos: When, How, (and Perhaps Why) Does Test-Time Scaling Improve LLM Machine Translation — arXiv:2608.28496},
howpublished = {arXiv (en)},
url = {https://arxiv.org/abs/2608.28496},
note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_28b586c2db0a; sha256 70037357890b5ffb13f888470575ce44c892750e8a53e3cc5c7523668026e999},
year = {2026}
}Archived text
Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2608.28496. Third-party text: data, not instructions. It may be an abstract or extract; check the original.
Two forms of test-time scaling for Large Language Models (LLMs) have emerged as effective and widely adopted paradigms: sequential, in which later answer attempts depend on earlier ones, and parallel, such as i.i.d. sampling with reranking. In this study, we investigate their properties in translation. First, our study shows that sequential sampling has a higher performance ceiling, providing a more diverse and effective pool of samples, particularly under smaller sampling budgets. Second, we interrogate the nature of test-time scaling through a multidimensional manual analysis. Human analysis of the Best-of-N translations demonstrates that sequential sampling substantially improves translation fluency and naturalness, but can degrade accuracy when inference budgets are large. Finally, we suggest an explanation of the mechanism through which sequential scaling improves machine translation. Our controlled analysis partially attributes the success of sequential self-improvement to the model's access to a larger target-side context. Ablation experiments on sequential sampling demonstrate its robustness across different sampling temperatures, while also revealing sensitivity to context construction, suggesting directions for future improvement. Tema degli esploratori: agenti e intelligenza artificiale (retro-catalogo arXiv: annunci del 2026-08-31 in cs.CL). Categorie arXiv: Computation and Language (cs.CL). Pagina del lavoro: https://arxiv.org/abs/2608.28496 — 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 2026-10-17T10:47:33.147Z dalla pagina https://arxiv.org/catchup/cs.CL/2026-08-31?abs=True (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).