mlsanityData to 14 Oct 2026

Active Learning by Acquiring Contrastive Examples

arXiv 2109.03764 ↗First posted 8 Sep 2021cs.CL cs.AI cs.LGcited by 156

Not on the map. Why?

Abstract (arXiv)

Common acquisition functions for active learning use either uncertainty or diversity sampling, aiming to select difficult and diverse data points from the pool of unlabeled data, respectively. In this work, leveraging the best of both worlds, we propose an acquisition function that opts for selecting textitcontrastive examples, i.e. data points that are similar in the model feature space and yet the model outputs maximally different predictive likelihoods. We compare our approach, CAL (Contrastive Active Learning), with a diverse set of acquisition functions in four natural language understanding tasks and seven datasets. Our experiments show that CAL performs consistently better or equal than the best performing baseline across all tasks, on both in-domain and out-of-domain data. We also conduct an extensive ablation study of our method and we further analyze all actively acquired datasets showing that CAL achieves a better trade-off between uncertainty and diversity compared to other strategies.

Key facts. Cited by 156 papers in the map (data through 14 Oct 2026). Builds on 31 of its 55 references that are in the map. First posted on arXiv 8 Sep 2021.

Builds on

31 of 55 references.

References · 31

  1. 2018BERT: Pre-training of Deep Bidirectional Transformers for Language Understandingcited by 40,399
  2. 2019PyTorch: An Imperative Style, High-Performance Deep Learning Librarycited by 20,305
  3. 2020A Simple Framework for Contrastive Learning of Visual Representationscited by 12,049
  4. 2018Representation Learning with Contrastive Predictive Codingcited by 7,777
  5. 2018GLUE: A Multi-Task Benchmark and Analysis Platform for Natural Language Understandingcited by 4,899
  6. 2015Dropout as a Bayesian Approximation: Representing Model Uncertainty in Deep Learningcited by 4,448
  7. 2013Distributed Representations of Words and Phrases and their Compositionalitycited by 3,474
  8. 2015Character-level Convolutional Networks for Text Classificationcited by 2,330
  9. 2017Deep Bayesian Active Learning with Image Datacited by 819
  10. 2019Deep Batch Active Learning by Diverse, Uncertain Gradient Lower Boundscited by 610
  11. 2019Learning the Difference that Makes a Difference with Counterfactually-Augmented Datacited by 439
  12. 2019BatchBALD: Efficient and Diverse Batch Acquisition for Deep Bayesian Active Learningcited by 432
19 more
  1. 2020Dataset Cartography: Mapping and Diagnosing Datasets with Training Dynamicscited by 425
  2. 2020Fine-Tuning Pretrained Language Models: Weight Initializations, Data Orders, and Early Stoppingcited by 384
  3. 2020Calibration of Pre-trained Transformerscited by 315
  4. 2020Pretrained Transformers Improve Out-of-Distribution Robustnesscited by 294
  5. 2020Supervised Contrastive Learning for Pre-trained Language Model Fine-tuningcited by 276
  6. 2018Adversarial Active Learning for Deep Networks: a Margin Based Approachcited by 158
  7. 2020Cold-start Active Learning through Self-supervised Language Modelingcited by 128
  8. 2017PubMed 200k RCT: a Dataset for Sequential Sentence Classification in Medical Abstractscited by 124
  9. 2017Deep Active Learning for Named Entity Recognitioncited by 108
  10. 2019Discriminative Active Learningcited by 98
  11. 2018Deep Bayesian Active Learning for Natural Language Processing: Results of a Large-Scale Empirical Studycited by 90
  12. 2019Diverse mini-batch Active Learningcited by 89
  13. 2017Deep Active Learning over the Long Tailcited by 84
  14. 2017Learning how to Active Learn: A Deep Reinforcement Learning Approachcited by 79
  15. 2017A Continuously Growing Dataset of Sentential Paraphrasescited by 72
  16. 2018Practical Obstacles to Deploying Active Learningcited by 68
  17. 2021Mind Your Outliers! Investigating the Negative Impact of Outliers on Active Learning for Visual Question Answeringcited by 62
  18. 2021Active Learning for Sequence Tagging with Deep Pre-trained Models and Bayesian Uncertainty Estimatescited by 39
  19. 2020Active Sentence Learning by Adversarial Uncertainty Sampling in Discrete Spacecited by 17

Built on by

Cited by year (2026 to 14 Oct)

Citing papers per year: 2021: 4; 2022: 19; 2023: 43; 2024: 41; 2025: 33; 2026: 16.

156 papers.

Sort:

By area:

  1. 2022Selective Annotation Makes Language Models Better Few-Shot Learnerscited by 243
  2. 2022DeepCore: A Comprehensive Library for Coreset Selection in Deep Learningcited by 147
  3. 2023Finding Support Examples for In-Context Learningcited by 104
  4. 2023InfoBatch: Lossless Training Speed Up by Unbiased Dynamic Data Pruningcited by 90
  5. 2023LLMaAA: Making Large Language Models as Active Annotatorscited by 79
  6. 2022Efficient Methods for Natural Language Processing: A Surveycited by 78
  7. 2023Large-scale Dataset Pruning with Dynamic Uncertaintycited by 63
  8. 2023Active Learning Principles for In-Context Learning with Large Language Modelscited by 54
  9. 2024Can Unconfident LLM Annotations Be Used for Confident Conclusions?cited by 49
  10. 2023Dataset Quantizationcited by 48
  11. 2023Uncertainty in Natural Language Processing: Sources, Quantification, and Applicationscited by 44
  12. 2022A Survey of Active Learning for Natural Language Processingcited by 40
38 more
  1. 2023Spanning Training Progress: Temporal Dual-Depth Scoring (TDDS) for Enhanced Dataset Pruningcited by 38
  2. 2023Which Examples to Annotate for In-Context Learning? Towards Effective and Efficient Selectioncited by 37
  3. 2023Loss-Curvature Matching for Dataset Selection and Condensationcited by 33
  4. 2025A Coreset Selection of Coreset Selection Literature: Introduction and Recent Advancescited by 28
  5. 2022A Rationale-Centric Framework for Human-in-the-loop Machine Learningcited by 27
  6. 2021Cartography Active Learningcited by 26
  7. 2023The Efficiency Spectrum of Large Language Models: An Algorithmic Surveycited by 26
  8. 2023Selective In-Context Data Augmentation for Intent Detection using Pointwise V-Informationcited by 24
  9. 2022ALLSH: Active Learning Guided by Local Sensitivity and Hardnesscited by 23
  10. 2022MEAL: Stable and Active Learning for Few-Shot Promptingcited by 20
  11. 2023You Only Condense Once: Two Rules for Pruning Condensed Datasetscited by 20
  12. 2024SHED: Shapley-Based Automated Dataset Refinement for Instruction Fine-Tuningcited by 19
  13. 2022Cold-Start Data Selection for Few-shot Language Model Fine-tuning: A Prompt-Based Uncertainty Propagation Approachcited by 16
  14. 2024The Power of LLM-Generated Synthetic Data for Stance Detection in Online Political Discussionscited by 16
  15. 2022Federated Few-Shot Learning for Mobile NLPcited by 15
  16. 2023Iterative Forward Tuning Boosts In-Context Learning in Language Modelscited by 15
  17. 2023Generalized Category Discovery with Large Language Models in the Loopcited by 15
  18. 2022PRBoost: Prompt-Based Rule Discovery and Boosting for Interactive Weakly-Supervised Learningcited by 14
  19. 2023ActiveGLAE: A Benchmark for Deep Active Learning with Transformerscited by 14
  20. 2023FreeAL: Towards Human-Free Active Learning in the Era of Large Language Modelscited by 14
  21. 2023More Samples or More Prompts? Exploring Effective In-Context Sampling for LLM Few-Shot Prompt Engineeringcited by 13
  22. 2024Dataset Quantization with Active Learning based Adaptive Samplingcited by 13
  23. 2024Post-hoc Probabilistic Vision-Language Modelscited by 13
  24. 2023D-CALM: A Dynamic Clustering-based Active Learning Approach for Mitigating Biascited by 12
  25. 2022Active Few-Shot Learning with FASLcited by 11
  26. 2023Data pruning and neural scaling laws: fundamental limitations of score-based algorithmscited by 11
  27. 2024Towards Robust and Efficient Cloud-Edge Elastic Model Adaptation via Selective Entropy Distillationcited by 11
  28. 2024ELFS: Label-Free Coreset Selection with Proxy Training Dynamicscited by 10
  29. 2024Active Learning for Robust and Representative LLM Generation in Safety-Critical Scenarioscited by 10
  30. 2021LMTurk: Few-Shot Learners as Crowdsourcing Workers in a Language-Model-as-a-Service Frameworkcited by 9
  31. 2023On the Limitations of Simulating Active Learningcited by 9
  32. 2023Towards Balanced Active Learning for Multimodal Classificationcited by 9
  33. 2023Towards Informative Few-Shot Prompt with Maximum Information Gain for In-Context Learningcited by 9
  34. 2024Beyond Efficiency: Molecular Data Pruning for Enhanced Generalizationcited by 9
  35. 2022FAMIE: A Fast Active Learning Framework for Multilingual Information Extractioncited by 8
  36. 2024Integrating Deep Metric Learning with Coreset for Active Learning in 3D Segmentationcited by 8
  37. 2024Unlocking Instructive In-Context Learning with Tabular Prompting for Relational Triple Extractioncited by 7
  38. 2024Evolution-aware VAriance (EVA) Coreset Selection for Medical Image Classificationcited by 7

50 most cited of 156 All 156 (JSON)