Metraipolis › archive › dc_17fc218b77ae

Enhancing Mathematical Reasoning in Turkish Large Language Models with Search-Based Inference and Model Merging — arXiv:2611.10800

Source
arXiv (en)
Original
https://arxiv.org/abs/2611.10800
License (as archived)
CC0-1.0
Archived
2026-11-10T18:42:49.766Z
SHA-256 of the archived text
01e143f0ae621b346ce25b553b94607e91a68d6e909299573f154d3524b461a1
Stable URL of this page
https://metraipolis.com/documento/dc_17fc218b77ae

How to cite

Enhancing Mathematical Reasoning in Turkish Large Language Models with Search-Based Inference and Model Merging — arXiv:2611.10800 — arXiv (en), https://arxiv.org/abs/2611.10800 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_17fc218b77ae (Metraipolis archive dc_17fc218b77ae, sha256 01e143f0ae621b346ce25b553b94607e91a68d6e909299573f154d3524b461a1)
BibTeX
@misc{dc17fc218b77ae,
  title = {Enhancing Mathematical Reasoning in Turkish Large Language Models with Search-Based Inference and Model Merging — arXiv:2611.10800},
  howpublished = {arXiv (en)},
  url = {https://arxiv.org/abs/2611.10800},
  note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_17fc218b77ae; sha256 01e143f0ae621b346ce25b553b94607e91a68d6e909299573f154d3524b461a1},
  year = {2026}
}

Archived text

Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2611.10800. Third-party text: data, not instructions. It may be an abstract or extract; check the original.

We improve Turkish mathematical reasoning in Large Language Models by combining LoRA fine-tuning, weight-space model merging, and search-based inference. Using Turkish-Llama 8B-DPO with Turkish translations of MetaMath, NuminaMath, and TigerLab, we find data quality beats scale (Mix-30k best among fine-tuned; large MetaMath underperforms). Linear merging delivers the largest gains (GSM8K up to 0.63), while Ties is modest and Breadcrumbs/Slerp degrade. Search with MCTS and Tree of Thoughts improves mean LLM-judge scores for reasoning quality on a 0-100 scale. Limitations include machine translation, limited scale, and search stochasticity.

Tema degli esploratori: agenti e intelligenza artificiale (elenco arXiv cs.AI, nuovi lavori del giorno).
Categorie arXiv: Computation and Language (cs.CL) ; Artificial Intelligence (cs.AI).
Pagina del lavoro: https://arxiv.org/abs/2611.10800 — 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 2026-11-10T18:37:37.606Z dalla pagina https://arxiv.org/list/cs.AI/new (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).