mlsanityData to 14 Oct 2026

Labeled TrustSet Guided: Batch Active Learning with Reinforcement Learning

arXiv 2604.12303 ↗First posted 14 Apr 2026cs.LGcited by 2

On the map: Deep reinforcement learning methods → Deep RL algorithms and representations

Abstract (arXiv)

Batch active learning (BAL) is a crucial technique for reducing labeling costs and improving data efficiency in training large-scale deep learning models. Traditional BAL methods often rely on metrics like Mahalanobis Distance to balance uncertainty and diversity when selecting data for annotation. However, these methods predominantly focus on the distribution of unlabeled data and fail to leverage feedback from labeled data or the model's performance. To address these limitations, we introduce TrustSet, a novel approach that selects the most informative data from the labeled dataset, ensuring a balanced class distribution to mitigate the long-tail problem. Unlike CoreSet, which focuses on maintaining the overall data distribution, TrustSet optimizes the model's performance by pruning redundant data and using label information to refine the selection process. To extend the benefits of TrustSet to the unlabeled pool, we propose a reinforcement learning (RL)-based sampling policy that approximates the selection of high-quality TrustSet candidates from the unlabeled data. Combining TrustSet and RL, we introduce the Batch Reinforcement Active Learning with TrustSet (BRAL-T) framework. BRAL-T achieves state-of-the-art results across 10 image classification benchmarks and 2 active fine-tuning tasks, demonstrating its effectiveness and efficiency in various domains.

Key facts. Cited by 2 papers in the map (data through 14 Oct 2026). Placed in Deep RL algorithms and representations, one of 15 directions in Deep reinforcement learning methods. Builds on 37 of its 52 references that are in the map. First posted on arXiv 14 Apr 2026.

Builds on

37 of 52 references.

References · 37

  1. 2021Emerging Properties in Self-Supervised Vision Transformerscited by 6,968
  2. 2020Training data-efficient image transformers & distillation through attentioncited by 4,586
  3. 2017Fashion-MNIST: a Novel Image Dataset for Benchmarking Machine Learning Algorithmscited by 4,349
  4. 2013Playing Atari with Deep Reinforcement Learningcited by 3,274
  5. 2016Identity Mappings in Deep Residual Networkscited by 1,965
  6. 2020WILDS: A Benchmark of in-the-Wild Distribution Shiftscited by 1,375
  7. 2019Distributionally Robust Neural Networks for Group Shifts: On the Importance of Regularization for Worst-Case Generalizationcited by 1,247
  8. 2017Deep Bayesian Active Learning with Image Datacited by 819
  9. 2018An Empirical Study of Example Forgetting during Deep Neural Network Learningcited by 678
  10. 2019Deep Batch Active Learning by Diverse, Uncertain Gradient Lower Boundscited by 610
  11. 2021Deep Learning on a Data Diet: Finding Important Examples Early in Trainingcited by 533
  12. 2019BatchBALD: Efficient and Diverse Batch Acquisition for Deep Bayesian Active Learningcited by 432
25 more
  1. 2019Learning Loss for Active Learningcited by 336
  2. 2019Variational Adversarial Active Learningcited by 284
  3. 2021Active Learning by Acquiring Contrastive Examplescited by 156
  4. 2020Cold-start Active Learning through Self-supervised Language Modelingcited by 128
  5. 2021Batch Active Learning at Scalecited by 127
  6. 2017Deep Active Learning for Named Entity Recognitioncited by 108
  7. 2019Deep Active Learning: Unified and Principled Method for Query and Trainingcited by 93
  8. 2022A Comparative Survey of Deep Active Learningcited by 89
  9. 2021Gone Fishing: Neural Active Learning with Fisher Embeddingscited by 88
  10. 2020Task-Aware Variational Adversarial Active Learningcited by 84
  11. 2021SIMILAR: Submodular Information Measures Based Active Learning In Realistic Scenarioscited by 84
  12. 2017Learning how to Active Learn: A Deep Reinforcement Learning Approachcited by 79
  13. 2023Data Pruning via Moving-one-Sample-outcited by 67
  14. 2016Coresets and Sketchescited by 63
  15. 2020Reinforced active learning for image segmentationcited by 56
  16. 2023Active Finetuning: Exploiting Annotation Budget in the Pretraining-Finetuning Paradigmcited by 36
  17. 2022GALAXY: Graph-based Active Learning at the Extremecited by 29
  18. 2022Plug and Play Active Learning for Object Detectioncited by 28
  19. 2023Entropic Open-set Active Learningcited by 23
  20. 2023Robust Data Pruning under Label Noise via Maximizing Re-labeling Accuracycited by 20
  21. 2023Algorithm Selection for Deep Active Learning with Imbalanced Datasetscited by 18
  22. 2022TiDAL: Learning Training Dynamics for Active Learningcited by 11
  23. 2021Multiple-criteria Based Active Learning with Fixed-size Determinantal Point Processescited by 9
  24. 2021MedSelect: Selective Labeling for Medical Image Classification Combining Meta-Learning with Deep Reinforcement Learningcited by 4
  25. 2024Compute-Efficient Active Learningcited by 2

Built on by

Cited by year (2026 to 14 Oct)

Citing papers per year: 2026: 2.

2 papers.

  1. 2026Active few-shot segmentation by reinforcing data selectioncited by 0
  2. 2026CAENL: Collapse-Based Active Neural Learning with Objective-Aligned Representation Controlcited by 0

Similar, not cited

Close in wording, no citation either way

  1. 2022TrustAL: Trustworthy Active Learning using Knowledge Distillation
  2. 2024Reinforcement Learning-Guided Semi-Supervised Learning
  3. 2026When Self-Belief Misleads: Active Label Acquisition for Reinforcement Learning with Verifiable Rewards
  4. 2025It's Not You, It's Clipping: A Soft Trust-Region via Probability Smoothing for LLM RL
  5. 2026Trust the Batch, On- or Off-Policy: Adaptive Policy Optimization for RL Post-Training
  6. 2022How to Leverage Unlabeled Data in Offline Reinforcement Learning
  7. 2026Boosting Reinforcement Learning with Verifiable Rewards via Randomly Selected Few-Shot Guidance
  8. 2026Rewards as Labels: Revisiting RLVR from a Classification Perspective