cd /news/large-language-models/scaling-domain-data-repetition-in-ll… · home topics large-language-models article
[ARTICLE · art-115802] src=arxiv.org ↗ pub= topic=large-language-models verified=true sentiment=· neutral

Scaling Domain Data Repetition in LLM Pretraining

A new arXiv preprint (2608.14071) finds that for large language models, the optimal number of times to repeat high-quality domain data during pretraining increases mildly with model size at a fixed tokens-per-parameter ratio, and is strongly negatively correlated with the domain's final validation loss. The authors, who studied practical LLM scaling where training-token budgets grow proportionally with model size, suggest that repetition counts tuned on smaller proxy models with the same TPP can provide a practical estimate for larger models.

read2 min views1 publishedAug 30, 2026
Scaling Domain Data Repetition in LLM Pretraining
Image: source
[Submitted on 14 Aug 2026]


[View PDF](/pdf/2608.14071)

Abstract:As large language models scale, their training-token budgets must also increase to maintain an appropriate tokens-per-parameter ratio ((\mathrm{TPP})). However, high-quality domain data is much harder to scale than general web data. As model size and the training-token budget increase, its fraction in the training mixture tends to decrease. Repeating the available high-quality data provides an effective way to counteract this dilution, but excessive repetition may lead to overfitting. We study this trade-off under practical LLM scaling, where the training-token budget grows proportionally with model size. For a fixed domain, we first find that, surprisingly at a fixed (\mathrm{TPP}), the optimal repetition count mildly increases with model size. Across different domains, we find that the optimal repetition count is strongly negatively correlated with the final validation loss of a domain: domains with lower loss can generally benefit from more repetitions. In contrast, the amount of unique domain data is only weakly related to the optimal repetition count. These findings suggest that repetition counts tuned on smaller proxy models with the same (\mathrm{TPP}) can provide a practical estimate for larger models.

References & Citations

...

Bibliographic Explorer

(What is the Explorer?) Connected Papers

(What is Connected Papers?) Litmaps

(What is Litmaps?) scite Smart Citations

(What are Smart Citations?)# Code, Data and Media Associated with this Article alphaXiv

(What is alphaXiv?) CatalyzeX Code Finder for Papers

(What is CatalyzeX?) DagsHub

(What is DagsHub?) Gotit.pub

(What is GotitPub?) Hugging Face

(What is Huggingface?) ScienceCast

(What is ScienceCast?)# Demos Influence Flower

(What are Influence Flowers?) CORE Recommender

(What is CORE?)# arXivLabs: experimental projects with community collaborators arXivLabs is a framework that allows collaborators to develop and share new arXiv features directly on our website.

Both individuals and organizations that work with arXivLabs have embraced and accepted our values of openness, community, excellence, and user data privacy. arXiv is committed to these values and only works with partners that adhere to them.

Have an idea for a project that will add value for arXiv's community? Learn more about arXivLabs.

── more in #large-language-models 4 stories · sorted by recency
── more on @arxiv 3 stories trending now
sponsored brought to you by zahid.host 4,200+ EU-deployed projects
reading about agents? ship yours in a single git push.

Run your AI side-project on zahid.host

EU-based hosting, git-push deploys, automatic HTTPS, no cold starts. Free tier with a custom domain — perfect for shipping the agent you just read about.

$git push zahid main
Live at https://your-agent.zahid.host
Get free account → Pricing
from €0/mo · no card required
LIVE [news/scaling-domain-data-…] indexed:0 read:2min 2026-08-30 ·