MIG: Automatic Data Selection for Instruction Tuning by Maximizing Information Gain in Semantic Space

• URL: http://arxiv.org/abs/2504.13835v1
• Date: Fri, 18 Apr 2025 17:59:46 GMT
• Title: MIG: Automatic Data Selection for Instruction Tuning by Maximizing Information Gain in Semantic Space
• Authors: Yicheng Chen, Yining Li, Kai Hu, Zerun Ma, Haochen Ye, Kai Chen,
• Abstract summary: Data quality and diversity are key to the construction of effective instruction-tuning datasets.<n>We introduce an efficient sampling method that selects data samples iteratively to textbfMaximize the textbfInformation textbfGain (MIG) in semantic space.
• Score: 12.583633720004118
• License: http://creativecommons.org/licenses/by/4.0/
• Abstract: Data quality and diversity are key to the construction of effective instruction-tuning datasets. % With the increasing availability of open-source instruction-tuning datasets, it is advantageous to automatically select high-quality and diverse subsets from a vast amount of data. % Existing methods typically prioritize instance quality and use heuristic rules to maintain diversity. % However, this absence of a comprehensive view of the entire collection often leads to suboptimal results. % Moreover, heuristic rules generally focus on distance or clustering within the embedding space, which fails to accurately capture the intent of complex instructions in the semantic space. % To bridge this gap, we propose a unified method for quantifying the information content of datasets. This method models the semantic space by constructing a label graph and quantifies diversity based on the distribution of information within the graph. % Based on such a measurement, we further introduce an efficient sampling method that selects data samples iteratively to \textbf{M}aximize the \textbf{I}nformation \textbf{G}ain (MIG) in semantic space. % Experiments on various datasets and base models demonstrate that MIG consistently outperforms state-of-the-art methods. % Notably, the model fine-tuned with 5\% Tulu3 data sampled by MIG achieves comparable performance to the official SFT model trained on the full dataset, with improvements of +5.73\% on AlpacaEval and +6.89\% on Wildbench.

Related papers

• Adaptive and Robust DBSCAN with Multi-agent Reinforcement Learning [53.527506374566485]
  We propose a novel Adaptive and Robust DBSCAN with Multi-agent Reinforcement Learning cluster framework, namely AR-DBSCAN.<n>We show that AR-DBSCAN not only improves clustering accuracy by up to 144.1% and 175.3% in the NMI and ARI metrics, respectively, but also is capable of robustly finding dominant parameters.
  arXiv  Detail & Related papers  (2025-05-07T11:37:23Z)
• QuaDMix: Quality-Diversity Balanced Data Selection for Efficient LLM Pretraining [12.872792775510172]
  We introduce a unified data selection framework called QuaDMix, which automatically optimize the data distribution for large language models pretraining.<n>Our experiments across diverse models and datasets demonstrate that QuaDMix achieves an average performance improvement of 7.2% across multiple benchmarks.
  arXiv  Detail & Related papers  (2025-04-23T08:36:50Z)
• MLLM-Selector: Necessity and Diversity-driven High-Value Data Selection for Enhanced Visual Instruction Tuning [69.7347209018861]
  We introduce MLLM-Selector, an automated approach that identifies valuable data for visual instruction tuning.<n>We calculate necessity scores for each sample in the VIT data pool to identify samples pivotal for enhancing model performance.<n>Our findings underscore the importance of mixing necessity and diversity in data choice, leading to the creation of MLLM-Selector.
  arXiv  Detail & Related papers  (2025-03-26T12:42:37Z)
• Diversity-Oriented Data Augmentation with Large Language Models [9.548912625579947]
  We propose a textbfunderlineDi-textbfunderlineoriented data textbfunderlineAugmentation framework (textbfDoAug)<n>Specifically, we utilize a diversity-oriented fine-tuning approach to train an LLM as a diverse paraphraser, which is capable of augmenting textual datasets by generating diversified paraphrases.<n>The results show that our fine-tuned LLM augmenter improves diversity while preserving label consistency, thereby enhancing the robustness and performance of downstream tasks.
  arXiv  Detail & Related papers  (2025-02-17T11:00:40Z)
• Semi-supervised Fine-tuning for Large Language Models [14.782756931646627]
  We introduce a semi-supervised fine-tuning(SemiFT) task and a framework named SemiEvol for LLM alignment.<n>For knowledge propagation, SemiEvol adopts a bi-level approach, propagating knowledge from labeled data to unlabeled data.<n>For knowledge selection, SemiEvol incorporates a collaborative learning mechanism, selecting higher-quality pseudo-response samples.
  arXiv  Detail & Related papers  (2024-10-17T16:59:46Z)
• A CLIP-Powered Framework for Robust and Generalizable Data Selection [51.46695086779598]
  Real-world datasets often contain redundant and noisy data, imposing a negative impact on training efficiency and model performance. Data selection has shown promise in identifying the most representative samples from the entire dataset. We propose a novel CLIP-powered data selection framework that leverages multimodal information for more robust and generalizable sample selection.
  arXiv  Detail & Related papers  (2024-10-15T03:00:58Z)
• Diversify and Conquer: Diversity-Centric Data Selection with Iterative Refinement [8.509688686402438]
  Finetuning large language models on instruction data is crucial for enhancing pre-trained knowledge and improving instruction-following capabilities. This work addresses the question: How can we determine the optimal subset of data for effective training? Our method employs k-means clustering to ensure the selected subset effectively represents the full dataset.
  arXiv  Detail & Related papers  (2024-09-17T17:25:31Z)
• RECOST: External Knowledge Guided Data-efficient Instruction Tuning [25.985023475991625]
  We argue that most current data-efficient instruction-tuning methods are highly dependent on the quality of the original instruction-tuning dataset. We propose a framework dubbed as textbfRECOST, which integrates external-knowledge-base re-ranking and diversity-consistent sampling into a single pipeline.
  arXiv  Detail & Related papers  (2024-02-27T09:47:36Z)
• How to Train Data-Efficient LLMs [56.41105687693619]
  We study data-efficient approaches for pre-training language models (LLMs) We find that Ask-LLM and Density sampling are the best methods in their respective categories. In our comparison of 19 samplers, involving hundreds of evaluation tasks and pre-training runs, we find that Ask-LLM and Density are the best methods in their respective categories.
  arXiv  Detail & Related papers  (2024-02-15T02:27:57Z)
• Consistency Regularization for Generalizable Source-free Domain Adaptation [62.654883736925456]
  Source-free domain adaptation (SFDA) aims to adapt a well-trained source model to an unlabelled target domain without accessing the source dataset. Existing SFDA methods ONLY assess their adapted models on the target training set, neglecting the data from unseen but identically distributed testing sets. We propose a consistency regularization framework to develop a more generalizable SFDA method.
  arXiv  Detail & Related papers  (2023-08-03T07:45:53Z)
• infoVerse: A Universal Framework for Dataset Characterization with Multidimensional Meta-information [68.76707843019886]
  infoVerse is a universal framework for dataset characterization. infoVerse captures multidimensional characteristics of datasets by incorporating various model-driven meta-information. In three real-world applications (data pruning, active learning, and data annotation), the samples chosen on infoVerse space consistently outperform strong baselines.
  arXiv  Detail & Related papers  (2023-05-30T18:12:48Z)
• Improving Semi-Supervised and Domain-Adaptive Semantic Segmentation with Self-Supervised Depth Estimation [94.16816278191477]
  We present a framework for semi-adaptive and domain-supervised semantic segmentation. It is enhanced by self-supervised monocular depth estimation trained only on unlabeled image sequences. We validate the proposed model on the Cityscapes dataset.
  arXiv  Detail & Related papers  (2021-08-28T01:33:38Z)
• Cross-Dataset Collaborative Learning for Semantic Segmentation [17.55660581677053]
  We present a simple, flexible, and general method for semantic segmentation, termed Cross-Dataset Collaborative Learning (CDCL) Given multiple labeled datasets, we aim to improve the generalization and discrimination of feature representations on each dataset. We conduct extensive evaluations on four diverse datasets, i.e., Cityscapes, BDD100K, CamVid, and COCO Stuff, with single-dataset and cross-dataset settings.
  arXiv  Detail & Related papers  (2021-03-21T09:59:47Z)

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.