mlsanityData to 14 Oct 2026

From Quantity to Quality: Boosting LLM Performance with Self-Guided Data Selection for Instruction Tuning

arXiv 2308.12032 ↗First posted 23 Aug 2023cs.CLcited by 307

Not on the map. Why?

Cited by 34 in LLM reasoning, but placed by its own abstract, which didn’t fit.

Abstract (arXiv)

In the realm of Large Language Models (LLMs), the balance between instruction data quality and quantity is a focal point. Recognizing this, we introduce a self-guided methodology for LLMs to autonomously discern and select cherry samples from open-source datasets, effectively minimizing manual curation and potential cost for instruction tuning an LLM. Our key innovation, the Instruction-Following Difficulty (IFD) metric, emerges as a pivotal metric to identify discrepancies between a model's expected responses and its intrinsic generation capability. Through the application of IFD, cherry samples can be pinpointed, leading to a marked uptick in model training efficiency. Empirical validations on datasets like Alpaca and WizardLM underpin our findings; with a mere 10% of original data input, our strategy showcases improved results. This synthesis of self-guided cherry-picking and the IFD metric signifies a transformative leap in the instruction tuning of LLMs, promising both efficiency and resource-conscious advancements. Codes, data, and models are available: https://github.com/tianyi-lab/CherryLLM

Key facts. Cited by 307 papers in the map (data through 14 Oct 2026). Builds on 40 of its 61 references that are in the map. First posted on arXiv 23 Aug 2023.

Builds on

40 of 61 references.

References · 40

  1. 2014Adam: A Method for Stochastic Optimizationcited by 37,240
  2. 2020Language Models are Few-Shot Learnerscited by 32,419
  3. 2023GPT-4 Technical Reportcited by 18,396
  4. 2022Training language models to follow instructions with human feedbackcited by 15,971
  5. 2023LLaMA: Open and Efficient Foundation Language Modelscited by 14,132
  6. 2023Llama 2: Open Foundation and Fine-Tuned Chat Modelscited by 13,028
  7. 2023Judging LLM-as-a-Judge with MT-Bench and Chatbot Arenacited by 8,626
  8. 2020Measuring Massive Multitask Language Understandingcited by 7,963
  9. 2018Think you have Solved Question Answering? Try ARC, the AI2 Reasoning Challengecited by 4,886
  10. 2019HellaSwag: Can a Machine Really Finish Your Sentence?cited by 4,100
  11. 2021Finetuned Language Models Are Zero-Shot Learnerscited by 3,780
  12. 2022FlashAttention: Fast and Memory-Efficient Exact Attention with IO-Awarenesscited by 3,673
28 more
  1. 2021TruthfulQA: Measuring How Models Mimic Human Falsehoodscited by 2,975
  2. 2022Self-Instruct: Aligning Language Models with Self-Generated Instructionscited by 2,724
  3. 2022BLOOM: A 176B-Parameter Open-Access Multilingual Language Modelcited by 1,984
  4. 2023A Survey on Evaluation of Large Language Modelscited by 1,242
  5. 2023WizardLM: Empowering large pre-trained language models to follow complex instructionscited by 1,125
  6. 2023LIMA: Less Is More for Alignmentcited by 1,100
  7. 2023Large Language Models are not Fair Evaluatorscited by 1,085
  8. 2021GLM: General Language Model Pretraining with Autoregressive Blank Infillingcited by 1,031
  9. 2022Super-NaturalInstructions: Generalization via Declarative Instructions on 1600+ NLP Taskscited by 868
  10. 2023Enhancing Chat Language Models by Scaling High-quality Instructional Conversationscited by 839
  11. 2023The RefinedWeb Dataset for Falcon LLM: Outperforming Curated Corpora with Web Data, and Web Data Onlycited by 763
  12. 2023AlpacaFarm: A Simulation Framework for Methods that Learn from Human Feedbackcited by 748
  13. 2023The Flan Collection: Designing Data and Methods for Effective Instruction Tuningcited by 739
  14. 2018An Empirical Study of Example Forgetting during Deep Neural Network Learningcited by 678
  15. 2021Deep Learning on a Data Diet: Finding Important Examples Early in Trainingcited by 533
  16. 2020UnifiedQA: Crossing Format Boundaries With a Single QA Systemcited by 476
  17. 2023Mitigating Hallucination in Large Multi-Modal Models via Robust Instruction Tuningcited by 445
  18. 2022Unnatural Instructions: Tuning Language Models with (Almost) No Human Laborcited by 397
  19. 2021Surface Form Competition: Why the Highest Probability Answer Isn't Always Rightcited by 263
  20. 2022Prioritized Training on Points that are Learnable, Worth Learning, and Not Yet Learntcited by 236
  21. 2021CrossFit: A Few-shot Learning Challenge for Cross-task Generalization in NLPcited by 142
  22. 2020Look at the First Sentence: Position Bias in Question Answeringcited by 84
  23. 2018On Coresets for Logistic Regressioncited by 67
  24. 2023HYTREL: Hypergraph-enhanced Tabular Data Representation Learningcited by 36
  25. 2016Sequence to Backward and Forward Sequences: A Content-Introducing Approach to Generative Short-Text Conversationcited by 35
  26. 2023Koala: An Index for Quantifying Overlaps with Pre-training Corporacited by 19
  27. 2019Unsupervised Context Rewriting for Open Domain Conversationcited by 14
  28. 2023Increasing Probability Mass on Answer Choices Does Not Always Improve Accuracycited by 11

Built on by

Cited by year (2026 to 14 Oct)

Citing papers per year: 2023: 11; 2024: 108; 2025: 119; 2026: 69.

307 papers.

Sort:

By area:

  1. 2025Kimi k1.5: Scaling Reinforcement Learning with LLMscited by 1,210
  2. 2023What Makes Good Data for Alignment? A Comprehensive Study of Automatic Data Selection in Instruction Tuningcited by 364
  3. 2024A Survey on Knowledge Distillation of Large Language Modelscited by 284
  4. 2024Continual Learning of Large Language Models: A Comprehensive Surveycited by 234
  5. 2025SoTA with Less: MCTS-Guided Sample Selection for Data-Efficient Visual Reasoning Self-Improvementcited by 145
  6. 2024Superfiltering: Weak-to-Strong Data Filtering for Fast Instruction-Tuningcited by 138
  7. 2024Rho-1: Not All Tokens Are What You Needcited by 133
  8. 2024OpenFedLLM: Training Large Language Models on Decentralized Private Data via Federated Learningcited by 128
  9. 2023MoDS: Model-oriented Data Selection for Instruction Tuningcited by 107
  10. 2024DART-Math: Difficulty-Aware Rejection Tuning for Mathematical Problem-Solvingcited by 95
  11. 2024A Survey on Self-Evolution of Large Language Modelscited by 91
  12. 2024A Survey on Data Selection for LLM Instruction Tuningcited by 77
38 more
  1. 2023Data Diversity Matters for Robust Instruction Tuningcited by 76
  2. 2024A Survey of Multimodal Large Language Model from A Data-centric Perspectivecited by 75
  3. 2024What is Wrong with Perplexity for Long-context Language Modeling?cited by 69
  4. 2025Pangu Embedded: An Efficient Dual-system LLM Reasoner with Metacognitioncited by 62
  5. 2024Selective Reflection-Tuning: Student-Selected Data Recycling for LLM Instruction-Tuningcited by 57
  6. 2024Clustering and Ranking: Diversity-preserved Instruction Selection through Expert-aligned Quality Estimationcited by 53
  7. 2025Intern-S1: A Scientific Multimodal Foundation Modelcited by 53
  8. 2025The Best Instruction-Tuning Data are Those That Fitcited by 52
  9. 2025Infinity Instruct: Scaling Instruction Selection and Synthesis to Enhance Language Modelscited by 50
  10. 2025DeSTA2.5-Audio: Toward General-Purpose Large Audio Language Model with Self-Generated Cross-Modal Alignmentcited by 50
  11. 2024Explaining Length Bias in LLM-Based Preference Evaluationscited by 49
  12. 2024Your Vision-Language Model Itself Is a Strong Filter: Towards High-Quality Instruction Tuning with Data Selectioncited by 48
  13. 2024Usable XAI: 10 Strategies Towards Exploiting Explainability in the LLM Eracited by 46
  14. 2025A Survey on the Optimization of Large Language Model-based Agentscited by 46
  15. 2023One-Shot Learning as Instruction Data Prospector for Large Language Modelscited by 43
  16. 2024Auto-RAG: Autonomous Retrieval-Augmented Generation for Large Language Modelscited by 43
  17. 2024Unveiling the Secret Recipe: A Guide For Supervised Fine-Tuning Small LLMscited by 40
  18. 2025Token Cleaning: Fine-Grained Data Selection for LLM Supervised Fine-Tuningcited by 39
  19. 2025Principled Data Selection for Alignment: The Hidden Risks of Difficult Examplescited by 37
  20. 2024Surveying the Effects of Quality, Diversity, and Complexity in Synthetic Data From Large Language Modelscited by 35
  21. 2025Large-Scale Data Selection for Instruction Tuningcited by 35
  22. 2023Rethinking the Instruction Quality: LIFT is What You Needcited by 34
  23. 2024Can LLMs Speak For Diverse People? Tuning LLMs via Debate to Generate Controllable Controversial Statementscited by 34
  24. 2024SelectIT: Selective Instruction Tuning for LLMs via Uncertainty-Aware Self-Reflectioncited by 34
  25. 2024Smaller Language Models are capable of selecting Instruction-Tuning Training Data for Larger Language Modelscited by 32
  26. 2024CoMM: Collaborative Multi-Agent, Multi-Reasoning-Path Prompting for Complex Problem Solvingcited by 31
  27. 2025ATLaS: Agent Tuning via Learning Critical Stepscited by 31
  28. 2023MUFFIN: Curating Multi-Faceted Instructions for Improving Instruction-Followingcited by 30
  29. 2024G-DIG: Towards Gradient-based Diverse and High-quality Instruction Data Selection for Machine Translationcited by 30
  30. 2024FedLLM-Bench: Realistic Benchmarks for Federated Learning of Large Language Modelscited by 30
  31. 2024TAGCOS: Task-agnostic Gradient Clustered Coreset Selection for Instruction Tuning Datacited by 30
  32. 2025MathFusion: Enhancing Mathematical Problem-solving of LLM through Instruction Fusioncited by 30
  33. 2024Diversity Measurement and Subset Selection for Instruction Tuning Datasetscited by 29
  34. 2024Unleashing the Power of Data Tsunami: A Comprehensive Survey on Data Assessment and Selection for Instruction Tuning of Language Modelscited by 29
  35. 2024Improving Data Efficiency via Curating LLM-Driven Rating Systemscited by 29
  36. 2024Automated Data Curation for Robust Language Model Fine-Tuningcited by 27
  37. 2024I-SHEEP: Self-Alignment of LLM from Scratch through an Iterative Self-Enhancement Paradigmcited by 27
  38. 2023Reflection-Tuning: Data Recycling Improves LLM Instruction-Tuningcited by 26

50 most cited of 307 All 307 (JSON)