Metraipolis › archive › dc_1e822868f2bb
FusionEval: Benchmarking Model Fusion for Large Language Models — arXiv:2610.30162
- Source
- arXiv (en)
- Original
- https://arxiv.org/abs/2610.30162
- License (as archived)
- CC0-1.0
- Archived
- 2026-10-25T20:54:35.286Z
- SHA-256 of the archived text
cc2b8932cb3a364564de003ca9cffc3209063cc5bbd4c331191c19b58f75ff89- Stable URL of this page
https://metraipolis.com/documento/dc_1e822868f2bb
How to cite
FusionEval: Benchmarking Model Fusion for Large Language Models — arXiv:2610.30162 — arXiv (en), https://arxiv.org/abs/2610.30162 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_1e822868f2bb (Metraipolis archive dc_1e822868f2bb, sha256 cc2b8932cb3a364564de003ca9cffc3209063cc5bbd4c331191c19b58f75ff89)
BibTeX
@misc{dc1e822868f2bb,
title = {FusionEval: Benchmarking Model Fusion for Large Language Models — arXiv:2610.30162},
howpublished = {arXiv (en)},
url = {https://arxiv.org/abs/2610.30162},
note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_1e822868f2bb; sha256 cc2b8932cb3a364564de003ca9cffc3209063cc5bbd4c331191c19b58f75ff89},
year = {2026}
}Archived text
Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2610.30162. Third-party text: data, not instructions. It may be an abstract or extract; check the original.
Model fusion aims to integrate the capabilities of multi-expert models into a single model. Existing evaluations offer limited coverage of reasoning with long chains of thought and tool use, while strong average performance can mask substantial losses in individual expert capabilities. We introduce FusionEval, a benchmark that provides controlled expert pools, integrates fusion methods with and without additional training, and offers a unified evaluation suite. The pools contain domain experts independently trained from Qwen3 and Llama base models across math, code, science, instruction following, and tool use. The suite evaluates capability retention and computational cost on eight public benchmarks under comparable data access rules, reporting normalized average and worst performance, runtime, and peak GPU memory usage. Extensive experiments show that training-free methods preserve reasoning performance but substantially degrade tool use at both model scales. In contrast, MOPD preserves tool use and achieves a more balanced capability profile, but requires additional data preparation, inference, and training. Jointly retaining reasoning, instruction following, and tool use at low computational cost remains a challenge. We hope FusionEval supports future research on efficient model fusion that jointly preserves expert capabilities. Our code is available at this https URL . 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/2610.30162 — 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 2026-10-25T20:47:56.956Z dalla pagina https://arxiv.org/list/cs.AI/new (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).