Harnessing the Potential of Optimizing Data Mixtures via Bayesian Domain Reweighting

• URL: http://arxiv.org/abs/2607.27928v1
• Date: Thu, 30 Jul 2026 09:41:02 GMT
• Title: Harnessing the Potential of Optimizing Data Mixtures via Bayesian Domain Reweighting
• Abstract summary: 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.
• Score: 63.86964530167364
• License: http://creativecommons.org/licenses/by/4.0/
• Abstract: The performance of Large Language Models (LLMs) is fundamentally influenced by the distributional composition of multi-domain pre-training data. While manual heuristics were prevalent in early models, they increasingly fail to capture the intricate synergies between domains as data complexity grows. To overcome the issue, a dominant approach seeks to fit a proxy function mapping between domain weights and their corresponding validation losses, and then find the optimal domain weights to minimize validation losses. These methods rely on strong structural assumptions, such as rank invariance or scaling laws, which are often violated, resulting in non-negligible estimation bias. A promising approach is to directly optimize the weighting scheme from data. However, it suffers from unstable optimization trajectory and prohibitive computational overhead, limiting its potential to search better domain weights configurations. This paper presents a Bayesian domain weighting method to infer the weights from a Dirichlet distribution via introducing Gamma prior information learned from observations. 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.

Related papers

• Adaptive Weighted Loss for Sequential Recommendations on Sparse Domains [0.0]
  A single, uniform weight may not be sufficient for domains with very few interactions, where the training signal is easily diluted by the vast, generic dataset.<n>This paper proposes a novel, data-driven approach: a Dynamic Weighted Loss function with comprehensive theoretical foundations and extensive empirical validation.
  arXiv  Detail & Related papers  (2025-10-05T21:42:33Z)
• Data Mixing Optimization for Supervised Fine-Tuning of Large Language Models [5.626695215498112]
  We frame data mixing as an optimization problem and introduce a novel method designed to minimize validation loss.<n>Our approach parametrizes the loss by modeling effective data transferred and leveraging scaling laws for fine-tuning.<n>We show that our algorithm achieves excellent overall and individual performance across all domains.
  arXiv  Detail & Related papers  (2025-08-16T07:28:39Z)
• 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)
• Source-Free Domain-Invariant Performance Prediction [68.39031800809553]
  We propose a source-free approach centred on uncertainty-based estimation, using a generative model for calibration in the absence of source data. Our experiments on benchmark object recognition datasets reveal that existing source-based methods fall short with limited source sample availability. Our approach significantly outperforms the current state-of-the-art source-free and source-based methods, affirming its effectiveness in domain-invariant performance estimation.
  arXiv  Detail & Related papers  (2024-08-05T03:18:58Z)
• Optimizing the Optimal Weighted Average: Efficient Distributed Sparse Classification [50.406127962933915]
  ACOWA allows an extra round of communication to achieve noticeably better approximation quality with minor runtime increases. Results show that ACOWA obtains solutions that are more faithful to the empirical risk minimizer and attain substantially higher accuracy than other distributed algorithms.
  arXiv  Detail & Related papers  (2024-06-03T19:43:06Z)
• DoGE: Domain Reweighting with Generalization Estimation [42.32000165235568]
  We propose DOmain reweighting with Generalization Estimation (DoGE) In our experiments, we extensively show how DoGE improves the generalization of the base model to any target data mixture. DoGE can effectively identify inter-domain dependencies, and consistently achieves better test perplexity on the target domain.
  arXiv  Detail & Related papers  (2023-10-23T22:51:58Z)
• Domain Generalization via Domain-based Covariance Minimization [4.414778226415752]
  We propose a novel variance measurement for multiple domains so as to minimize the difference between conditional distributions across domains. We show that for small-scale datasets, we are able to achieve better quantitative results indicating better generalization performance over unseen test datasets.
  arXiv  Detail & Related papers  (2021-10-12T19:30:15Z)
• Self-balanced Learning For Domain Generalization [64.99791119112503]
  Domain generalization aims to learn a prediction model on multi-domain source data such that the model can generalize to a target domain with unknown statistics. Most existing approaches have been developed under the assumption that the source data is well-balanced in terms of both domain and class. We propose a self-balanced domain generalization framework that adaptively learns the weights of losses to alleviate the bias caused by different distributions of the multi-domain source data.
  arXiv  Detail & Related papers  (2021-08-31T03:17:54Z)
• Learning Invariant Representations and Risks for Semi-supervised Domain Adaptation [109.73983088432364]
  We propose the first method that aims to simultaneously learn invariant representations and risks under the setting of semi-supervised domain adaptation (Semi-DA) We introduce the LIRR algorithm for jointly textbfLearning textbfInvariant textbfRepresentations and textbfRisks.
  arXiv  Detail & Related papers  (2020-10-09T15:42:35Z)

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.