mlsanityData to 14 Oct 2026

FIND: Human-in-the-Loop Debugging Deep Text Classifiers

arXiv 2010.04987 ↗First posted 10 Oct 2020cs.CL cs.HC cs.LGcited by 40

Not on the map. Why?

Abstract (arXiv)

Since obtaining a perfect training dataset (i.e., a dataset which is considerably large, unbiased, and well-representative of unseen cases) is hardly possible, many real-world text classifiers are trained on the available, yet imperfect, datasets. These classifiers are thus likely to have undesirable properties. For instance, they may have biases against some sub-populations or may not work effectively in the wild due to overfitting. In this paper, we propose FIND -- a framework which enables humans to debug deep learning text classifiers by disabling irrelevant hidden features. Experiments show that by using FIND, humans can improve CNN text classifiers which were trained under different types of imperfect datasets (including datasets with biases and datasets with dissimilar train-test distributions).

Key facts. Cited by 40 papers in the map (data through 14 Oct 2026). First posted on arXiv 10 Oct 2020.

Builds on

No reference list yet.

Built on by

Cited by year

Citing papers per year: 2021: 11; 2022: 13; 2023: 4; 2024: 9; 2025: 3.

40 papers.

Sort:

By area:

  1. 2022Post-hoc Concept Bottleneck Modelscited by 293
  2. 2021Explanation-Based Human Debugging of NLP Models: A Surveycited by 53
  3. 2022Leveraging Explanations in Interactive Machine Learning: An Overviewcited by 53
  4. 2022Concept-level Debugging of Part-Prototype Networkscited by 52
  5. 2021Interactive Label Cleaning with Example-based Explanationscited by 49
  6. 2021Refining Language Models with Compositional Explanationscited by 27
  7. 2022A Rationale-Centric Framework for Human-in-the-loop Machine Learningcited by 27
  8. 2023Interactive Explanations by Conflict Resolution via Argumentative Exchangescited by 24
  9. 2022Discovering Salient Neurons in Deep NLP Modelscited by 21
  10. 2021Jointly Attacking Graph Neural Network and its Explanationscited by 18
  11. 2024Position: Towards Bidirectional Human-AI Alignmentcited by 18
  12. 2025If Concept Bottlenecks are the Question, are Foundation Models the Answer?cited by 15
28 more
  1. 2021Toward Learning Human-aligned Cross-domain Robust Models by Countering Misaligned Featurescited by 14
  2. 2022Self-explaining deep models with logic rule reasoningcited by 13
  3. 2024Contestable AI needs Computational Argumentationcited by 13
  4. 2025Shortcuts and Identifiability in Concept-based Models from a Neuro-Symbolic Lenscited by 13
  5. 2024CREW: Facilitating Human-AI Teaming Researchcited by 11
  6. 2021Towards Benchmarking the Utility of Explanations for Model Debuggingcited by 10
  7. 2022XMD: An End-to-End Framework for Interactive Explanation-Based Debugging of NLP Modelscited by 8
  8. 2024A Survey on Human-AI Collaboration with Large Foundation Modelscited by 7
  9. 2021GrASP: A Library for Extracting and Exploring Human-Interpretable Textual Patternscited by 6
  10. 2022Explainable Patterns for Distinction and Prediction of Moral Judgement on Redditcited by 6
  11. 2022Causal Discovery and Knowledge Injection for Contestable Neural Networks (with Appendices)cited by 6
  12. 2022ER-Test: Evaluating Explanation Regularization Methods for Language Modelscited by 6
  13. 2023May I Ask a Follow-up Question? Understanding the Benefits of Conversations in Neural Network Explainabilitycited by 5
  14. 2021Toward a Unified Framework for Debugging Concept-based Modelscited by 4
  15. 2023Accelerating the Global Aggregation of Local Explanationscited by 4
  16. 2021Explaining the Road Not Takencited by 3
  17. 2024Test-Time Fairness and Robustness in Large Language Modelscited by 3
  18. 2022HumanAL: Calibrating Human Matching Beyond a Single Taskcited by 2
  19. 2022Understanding Text Classification Data and Models Using Aggregated Input Saliencecited by 2
  20. 2023Dual-Teacher De-biasing Distillation Framework for Multi-domain Fake News Detectioncited by 2
  21. 2024Joint Universal Adversarial Perturbations with Interpretationscited by 2
  22. 2021Bandits for Learning to Explain from Explanationscited by 1
  23. 2022Argumentative Explanations for Pattern-Based Text Classifierscited by 1
  24. 2024InfFeed: Influence Functions as a Feedback to Improve the Performance of Subjective Taskscited by 1
  25. 2024Advancing Interactive Explainable AI via Belief Change Theorycited by 1
  26. 2021A Symmetric Loss Perspective of Reliable Machine Learningcited by 0
  27. 2024Visual Analytics for Fine-grained Text Classification Models and Datasetscited by 0
  28. 2025Human Cognitive Biases in Explanation-Based Interaction: The Case of Within and Between Session Order Effectcited by 0