Lune

ICML2025Top-tier venue

The Best of Both Worlds: Bridging Quality and Diversity in Data Selection with Bipartite Graph

Minghao Wu, Thuy-Trang Vu, Lizhen Qu, Gholamreza Haffari

Abstract

The performance of large language models (LLMs) is strongly influenced by the quality and diversity of data used during supervised finetuning (SFT). However, current data selection methods often prioritize one aspect over the other, resulting in suboptimal training outcomes. To address this, we formulate data selection as a set cover problem and present GRAPHFILTER, a novel approach that balances both quality and diversity in data selection. GRAPHFILTER models the dataset as a bipartite graph connecting sentences to their constituent n-grams, then employs a priority function that combines quality and diversity metrics multiplicatively. GRAPHFILTER iteratively selects sentences with the highest priority, removes covered n-grams from the bipartite graph, and recomputes priorities to reflect the changing data landscape. We validate GRAPH-FILTER using three model backbones across six widely-used benchmarks, demonstrating that it outperforms nine existing baselines in both model performance and computational efficiency. Further analysis shows that our design choices lead to more effective subset selection, underscores the value of instruction diversity, and provides insights into how quality and diversity interact with different subset sizes.

Ask Lune

Know this paper in minutes, not an afternoon.

Lune can read every page of it and the top-tier papers that cite it, and gives you its claims, its evidence and its limits, each one cited to the paper it comes from. No skimming, no AI slop.

4 top-tier papers have cited it since. Know what they found before your reviewers do.

Questions to start from

Lune can draw on

the full paper · the top-tier papers that cite it · related work

Get the answer

Free to start. No credit card required.

Lineage

  • RICo: Refined In-Context Contribution for Automatic Instruction-Tuning Data Selection cites Measuring Massive Multitask Language Understanding
  • RICo: Refined In-Context Contribution for Automatic Instruction-Tuning Data Selection cites Multitask Prompted Training Enables Zero-Shot Task Generalization
  • RICo: Refined In-Context Contribution for Automatic Instruction-Tuning Data Selection cites Self-Instruct: Aligning Language Models with Self-Generated Instructions
  • HBO: Hierarchical Balancing Optimization for Fine-Tuning Large Language Models cites Measuring Massive Multitask Language Understanding
  • Unifying and Optimizing Data Values for Selection via Sequential Decision-Making cites Measuring Massive Multitask Language Understanding
  • LIMA: Less Is More for Alignment cites Multitask Prompted Training Enables Zero-Shot Task Generalization
  • LIMA: Less Is More for Alignment cites Training language models to follow instructions with human feedback
  • Training language models to follow instructions with human feedback cites Multitask Prompted Training Enables Zero-Shot Task Generalization
  • Self-Instruct: Aligning Language Models with Self-Generated Instructions cites Training language models to follow instructions with human feedback
Builds onCited byLeft to right by year; the more cited, the higher.

Cited by top-tier papers4

  1. ReasonFlux-PRM: Trajectory-Aware PRMs for Long Chain-of-Thought Reasoning in LLMsJiaru Zou, Ling Yang, Jingwen Gu, Jiahao Qiu et al.NeurIPS 2025 · 51 citations
  2. HBO: Hierarchical Balancing Optimization for Fine-Tuning Large Language ModelsWeixuan Wang, Minghao Wu, Barry Haddow, Alexandra BirchICLR 2026 · 2 citations
  3. Unifying and Optimizing Data Values for Selection via Sequential Decision-MakingFrank Hongliang Chi, Qiong Wu, Zhengyi Zhou, Jonathan Li et al.ICML 2026 · 1 citation
  4. RICo: Refined In-Context Contribution for Automatic Instruction-Tuning Data SelectionYixin Yang, Qingxiu Dong, Linli Yao, Fangwei Zhu et al.AAAI 2026

Builds on15

  1. Training language models to follow instructions with human feedbackLong Ouyang, Jeffrey Wu, Xu Jiang, Diogo Almeida et al.NeurIPS 2022 · 24,707 citations
  2. Measuring Massive Multitask Language UnderstandingDan Hendrycks, Collin Burns, Steven Basart, Andy Zou et al.ICLR 2021 · 7,905 citations
  3. Multitask Prompted Training Enables Zero-Shot Task GeneralizationVictor Sanh, Albert Webson, Colin Raffel, Stephen H. Bach et al.ICLR 2022 · 1,976 citations
  4. LIMA: Less Is More for AlignmentChunting Zhou, Pengfei Liu, Puxin Xu, Srinivasan Iyer et al.NeurIPS 2023 · 1,486 citations
  5. Self-Instruct: Aligning Language Models with Self-Generated InstructionsYizhong Wang, Yeganeh Kordi, Swaroop Mishra, Alisa Liu et al.ACL 2023 · 540 citations

Related papers

  1. Your Vision-Language Model Itself Is a Strong Filter: Towards High-Quality Instruction Tuning with Data SelectionRuibo Chen, Yihan Wu, Lichang Chen, Guodong Liu et al.ACL Findings 2024 · 5 citations
  2. Rethinking Data Selection at Scale: Random Selection is Almost All You NeedTingyu Xia, Bowen Yu, Kai Dang, An Yang et al.EMNLP Findings 2025 · 4 citations
  3. G-DIG: Towards Gradient-based DIverse and hiGh-quality Instruction Data Selection for Machine TranslationXingyuan Pan, Luyang Huang, Liyan Kang, Zhicheng Liu et al.ACL 2024
  4. Superfiltering: Weak-to-Strong Data Filtering for Fast Instruction-TuningMing Li, Yong Zhang, Shwai He, Zhitao Li et al.ACL 2024 · 16 citations
  5. MergeIT: From Selection to Merging for Efficient Instruction TuningHongyi Cai, Yuqian Fu, Hongming Fu, Bo ZhaoACL Findings 2026
Dusk over the sea between two cliffs drawn in fine vertical lines