Domain-Aware Scaling Laws Uncover Data Synergy

• URL: http://arxiv.org/abs/2607.11052v1
• Date: Mon, 13 Jul 2026 03:31:03 GMT
• Title: Domain-Aware Scaling Laws Uncover Data Synergy
• Abstract summary: We formalize and quantify data synergy in language model pretraining.<n>Our framework improves predictive accuracy over domain-agnostic scaling laws.<n>We validate these estimates by training models on predicted optimal and predicted anti-optimal mixtures.
• Score: 22.97719264150389
• License: http://creativecommons.org/licenses/by/4.0/
• Abstract: Machine learning progress is often attributed to scaling model size and dataset volume, yet the composition of data can be just as consequential. Empirical findings repeatedly show that combining datasets from different domains yields nontrivial interactions. For instance, adding code improves mathematical reasoning, while certain mixtures introduce interference that reduces model performance. We refer to these effects collectively as data synergy, where the contribution of multiple domains exceeds or falls short of the sum of their isolated contributions. In this work, we formalize and quantify data synergy in language model pretraining. Leveraging observational variation across open-weight LLMs with diverse pretraining mixtures, we estimate both direct domain-to-benchmark synergy (how one domain contributes to performance on another) and a second-order domain-domain synergy (capabilities that require co-occurrence of multiple domains). Our framework improves predictive accuracy over domain-agnostic scaling laws and recovers stable synergy estimates. We validate these estimates by training models on predicted optimal and predicted anti-optimal mixtures and confirm that our synergy estimates correctly predict performance rankings.

Related papers

• Harnessing the Potential of Optimizing Data Mixtures via Bayesian Domain Reweighting [63.86964530167364]
  This paper presents a Bayesian domain weighting method to infer the weights from a Dirichlet distribution via introducing Gamma prior information learned from observations.<n> Experimental results demonstrate that proposed method could achieve stable and efficient domain weights learning, and identifies optimal mixtures while consuming substantially less data than search-based function-fitting methods, revitalizing optimization-based domain weighting for large-scale applications.
  arXiv  Detail & Related papers  (2026-07-30T09:41:02Z)
• Explaining Data Mixing Scaling Laws [2.1378946778058796]
  We propose a unified framework to explain the underlying mechanics of data mixing.<n>We identify two key factors that govern the domain losses of models trained on different data mixtures.<n>Our model successfully extrapolates across scales, predicting highly effective mixtures for large, unseen scales.
  arXiv  Detail & Related papers  (2026-06-06T13:31:38Z)
• Data Mixing Agent: Learning to Re-weight Domains for Continual Pre-training [30.915768238214653]
  Data Mixing Agent is an end-to-end framework that learns to re-weight domains.<n>It generalizes well across unseen source fields, target models, and domain spaces without retraining.
  arXiv  Detail & Related papers  (2025-07-21T14:01:54Z)
• DIDS: Domain Impact-aware Data Sampling for Large Language Model Training [61.10643823069603]
  We present Domain Impact-aware Data Sampling (DIDS) for large language models.<n>DIDS group training data based on learning effects, where a proxy language model and dimensionality reduction are employed.<n>It achieves 3.4% higher average performance while maintaining comparable training efficiency.
  arXiv  Detail & Related papers  (2025-04-17T13:09:38Z)
• Collaborative Value Function Estimation Under Model Mismatch: A Federated Temporal Difference Analysis [55.13545823385091]
  Federated reinforcement learning (FedRL) enables collaborative learning while preserving data privacy by preventing direct data exchange between agents.<n>In real-world applications, each agent may experience slightly different transition dynamics, leading to inherent model mismatches.<n>We show that even moderate levels of information sharing significantly mitigate environment-specific errors.
  arXiv  Detail & Related papers  (2025-03-21T18:06:28Z)
• Towards Theoretical Understandings of Self-Consuming Generative Models [56.84592466204185]
  This paper tackles the emerging challenge of training generative models within a self-consuming loop. We construct a theoretical framework to rigorously evaluate how this training procedure impacts the data distributions learned by future models. We present results for kernel density estimation, delivering nuanced insights such as the impact of mixed data training on error propagation.
  arXiv  Detail & Related papers  (2024-02-19T02:08:09Z)
• Structured Radial Basis Function Network: Modelling Diversity for Multiple Hypotheses Prediction [51.82628081279621]
  Multi-modal regression is important in forecasting nonstationary processes or with a complex mixture of distributions. A Structured Radial Basis Function Network is presented as an ensemble of multiple hypotheses predictors for regression problems. It is proved that this structured model can efficiently interpolate this tessellation and approximate the multiple hypotheses target distribution.
  arXiv  Detail & Related papers  (2023-09-02T01:27:53Z)
• Transformer Based Multi-Source Domain Adaptation [53.24606510691877]
  In practical machine learning settings, the data on which a model must make predictions often come from a different distribution than the data it was trained on. Here, we investigate the problem of unsupervised multi-source domain adaptation, where a model is trained on labelled data from multiple source domains and must make predictions on a domain for which no labelled data has been seen. We show that the predictions of large pretrained transformer based domain experts are highly homogenous, making it challenging to learn effective functions for mixing their predictions.
  arXiv  Detail & Related papers  (2020-09-16T16:56:23Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.