mlsanityData to 7 Oct 2026

Exploring Empty Spaces: Human-in-the-Loop Data Augmentation

arXiv 2410.01088 ↗First posted 1 Oct 2024cs.HC cs.CL cs.LGcited by 6

On the map: LLM safety and security → Machine unlearning

Abstract (arXiv)

Data augmentation is crucial to make machine learning models more robust and safe. However, augmenting data can be challenging as it requires generating diverse data points to rigorously evaluate model behavior on edge cases and mitigate potential harms. Creating high-quality augmentations that cover these "unknown unknowns" is a time- and creativity-intensive task. In this work, we introduce Amplio, an interactive tool to help practitioners navigate "unknown unknowns" in unstructured text datasets and improve data diversity by systematically identifying empty data spaces to explore. Amplio includes three human-in-the-loop data augmentation techniques: Augment With Concepts, Augment by Interpolation, and Augment with Large Language Model. In a user study with 18 professional red teamers, we demonstrate the utility of our augmentation methods in helping generate high-quality, diverse, and relevant model safety prompts. We find that Amplio enabled red teamers to augment data quickly and creatively, highlighting the transformative potential of interactive augmentation workflows.

Key facts. Cited by 6 papers in the map (data through 7 Oct 2026). Placed in Machine unlearning, one of 12 directions in LLM safety and security. Builds on 60 of its 119 references that are in the map. First posted on arXiv 1 Oct 2024.

Builds on

60 of 119 references.

References · 60

  1. 2019Sentence-BERT: Sentence Embeddings using Siamese BERT-Networkscited by 9,049
  2. 2017mixup: Beyond Empirical Risk Minimizationcited by 5,155
  3. 2022Training a Helpful and Harmless Assistant with Reinforcement Learning from Human Feedbackcited by 3,634
  4. 2018UMAP: Uniform Manifold Approximation and Projection for Dimension Reductioncited by 2,841
  5. 2019RandAugment: Practical automated data augmentation with a reduced search spacecited by 2,261
  6. 2015A Diversity-Promoting Objective Function for Neural Conversation Modelscited by 1,068
  7. 2019Unsupervised Data Augmentation for Consistency Trainingcited by 1,057
  8. 2024WildChat: 1M ChatGPT Interaction Logs in the Wildcited by 598
  9. 2018Texygen: A Benchmarking Platform for Text Generation Modelscited by 484
  10. 2022The Vendi Score: A Diversity Evaluation Metric for Machine Learningcited by 299
  11. 2017Toward Controlled Generation of Textcited by 271
  12. 2021Polyjuice: Generating Counterfactuals for Explaining, Evaluating, and Improving Modelscited by 219
38 more
  1. 2019The What-If Tool: Interactive Probing of Machine Learning Modelscited by 207
  2. 2024WildTeaming at Scale: From In-the-Wild Jailbreaks to (Adversarially) Safer Language Modelscited by 206
  3. 2021Data Augmentation Can Improve Robustnesscited by 193
  4. 2023Text Embeddings Reveal (Almost) As Much As Textcited by 189
  5. 2024Rainbow Teaming: Open-Ended Generation of Diverse Adversarial Promptscited by 189
  6. 2020MixText: Linguistically-Informed Interpolation of Hidden Space for Semi-Supervised Text Classificationcited by 186
  7. 2024Improving Dictionary Learning with Gated Sparse Autoencoderscited by 171
  8. 2020The Language Interpretability Tool: Extensible, Interactive Visualizations and Analysis for NLP Modelscited by 125
  9. 2021A Survey on Data Augmentation for Text Classificationcited by 111
  10. 2023ChainForge: A Visual Toolkit for Prompt Engineering and LLM Hypothesis Testingcited by 88
  11. 2024Red-Teaming for Generative AI: Silver Bullet or Security Theater?cited by 81
  12. 2019Summit: Scaling Deep Learning Interpretability by Visualizing Activation and Attribution Summarizationscited by 72
  13. 2018Seq2Seq-Vis: A Visual Debugging Tool for Sequence-to-Sequence Modelscited by 69
  14. 2023Promptify: Text-to-Image Generation through Interactive Prompt Exploration with Large Language Modelscited by 67
  15. 2019FairSight: Visual Analytics for Fairness in Decision Makingcited by 59
  16. 2023PromptMagician: Interactive Prompt Engineering for Text-to-Image Creationcited by 58
  17. 2020CNN Explainer: Learning Convolutional Neural Networks with Interactive Visualizationcited by 57
  18. 2019A Closer Look At Feature Space Data Augmentation For Few-Shot Intent Classificationcited by 45
  19. 2016Embedding Projector: Interactive Visualization and Interpretation of Embeddingscited by 43
  20. 2024DataDreamer: A Tool for Synthetic Data Generation and Reproducible LLM Workflowscited by 43
  21. 2020GenAug: Data Augmentation for Finetuning Text Generatorscited by 41
  22. 2020CoDA: Contrast-enhanced and Diversity-promoting Data Augmentation for Natural Language Understandingcited by 41
  23. 2022Unleashing the Power of Graph Data Augmentation on Covariate Distribution Shiftcited by 38
  24. 2023PromptPaint: Steering Text-to-Image Generation Through Paint Medium-like Interactionscited by 34
  25. 2020Towards Accuracy-Fairness Paradox: Adversarial Example-based Data Augmentation for Visual Debiasingcited by 32
  26. 2024LLM Comparator: Visual Analytics for Side-by-Side Evaluation of Large Language Modelscited by 29
  27. 2024Farsight: Fostering Responsible AI Awareness During AI Application Prototypingcited by 28
  28. 2024Supporting Sensemaking of Large Language Model Outputs at Scalecited by 26
  29. 2023Mixture of Soft Prompts for Controllable Data Generationcited by 23
  30. 2023Prompting for Discovery: Flexible Sense-Making for AI Art-Making with Dreamsheetscited by 22
  31. 2020Diversity, Density, and Homogeneity: Quantitative Characteristic Metrics for Text Collectionscited by 21
  32. 2020Narrative Interpolation for Generating and Understanding Storiescited by 19
  33. 2020Tell Me How to Ask Again: Question Data Augmentation with Controllable Rewriting in Continuous Spacecited by 19
  34. 2024GhostWriter: Augmenting Collaborative Human-AI Writing Experiences Through Personalization and Agencycited by 18
  35. 2023Designing LLM Chains by Adapting Techniques from Crowdsourcing Workflowscited by 17
  36. 2017Paying Attention to Descriptions Generated by Image Captioning Modelscited by 12
  37. 2022Erato: Cooperative Data Story Editing via Fact Interpolationcited by 11
  38. 2024PrompTHis: Visualizing the Process and Influence of Prompt Editing during Text-to-Image Creationcited by 11

50 most cited of 60 All 60 (JSON)

Built on by

Cited by year (2026 to 7 Oct)

Citing papers per year: 2025: 3; 2026: 3.

6 papers.

  1. 2025Culture is Everywhere: A Call for Intentionally Cultural Evaluationcited by 10
  2. 2025Generate, Evaluate, Iterate: Synthetic Data for Human-in-the-Loop Refinement of LLM Judgescited by 2
  3. 2025Data-Prompt Co-Evolution: Growing Test Sets to Refine LLM Behaviorcited by 1
  4. 2026The Consensus Trap: Dissecting Subjectivity and the "Ground Truth" Illusion in Data Annotationcited by 1
  5. 2026ctELM: Decoding and Manipulating Embeddings of Clinical Trials with Embedding Language Modelscited by 0
  6. 2026LLM-Guided Semantic Bootstrapping for Interpretable Text Classification with Tsetlin Machinescited by 0