Metraipolis › archive › dc_8922a7fd78ac
Can We Optimize the Performance-Carbon Emission Break-Even Point?: The Quest for Greener LLMs — arXiv:2608.08744
- Source
- arXiv (en)
- Original
- https://arxiv.org/abs/2608.08744
- License (as archived)
- CC0-1.0
- Archived
- 2026-10-19T08:50:04.631Z
- SHA-256 of the archived text
9849a5de876c33cde0f94dffa82b310f7489fdda938a405df93516011f70e06c- Stable URL of this page
https://metraipolis.com/documento/dc_8922a7fd78ac
How to cite
Can We Optimize the Performance-Carbon Emission Break-Even Point?: The Quest for Greener LLMs — arXiv:2608.08744 — arXiv (en), https://arxiv.org/abs/2608.08744 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_8922a7fd78ac (Metraipolis archive dc_8922a7fd78ac, sha256 9849a5de876c33cde0f94dffa82b310f7489fdda938a405df93516011f70e06c)
BibTeX
@misc{dc8922a7fd78ac,
title = {Can We Optimize the Performance-Carbon Emission Break-Even Point?: The Quest for Greener LLMs — arXiv:2608.08744},
howpublished = {arXiv (en)},
url = {https://arxiv.org/abs/2608.08744},
note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_8922a7fd78ac; sha256 9849a5de876c33cde0f94dffa82b310f7489fdda938a405df93516011f70e06c},
year = {2026}
}Archived text
Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2608.08744. Third-party text: data, not instructions. It may be an abstract or extract; check the original.
The carbon footprint of any deployed Large Language Model (LLM) accumulates during inference, where repeated use of the model substantially exceeds the one-time cost of fine-tuning. Yet most efficiency interventions target either pre-training scale or post-hoc compression. We ask whether folding a calibrated, differentiable energy surrogate into the fine-tuning objective can produce inference behavior that gains task accuracy at zero or near-zero carbon cost, a break-even configuration. We propose a joint loss mechanism with a per-model carbon-emission parameter, a linear surrogate over parameter norm, FLOP proxy, and a memory proxy, fit from on-hardware energy profiling. We fine-tune three architecturally distinct families: Gemma-2 2B, Llama-3.1 8B, and Qwen-2.5 14B, and evaluate inference F1 and CO$_2$ emissions on three MMLU subjects: abstract algebra, philosophy, and formal logic. We discover from several outcomes that the carbon term behaves as either harmful interference or beneficial regularization depending on the task structure. We position calibrated carbon-aware fine-tuning as a lightweight, drop-in regularizer with a non-empty but model and task-dependent break-even region. This is an ongoing work, and we will release our codebase soon. Tema degli esploratori: agenti e intelligenza artificiale (retro-catalogo arXiv: annunci del 2026-08-11 in cs.CL). Categorie arXiv: Computation and Language (cs.CL) ; Artificial Intelligence (cs.AI); Computational Engineering, Finance, and Science (cs.CE); Emerging Technologies (cs.ET); Machine Learning (cs.LG). Pagina del lavoro: https://arxiv.org/abs/2608.08744 — 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 2026-10-19T08:49:41.868Z dalla pagina https://arxiv.org/catchup/cs.CL/2026-08-11?abs=True (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).