mlsanityData to 7 Oct 2026

Enhancing the QA Model through a Multi-domain Debiasing Framework

arXiv 2601.11581 ↗First posted 1 Jan 2026cs.CL cs.AIcited by 0

Not on the map. Why?

Abstract (arXiv)

Question-answering (QA) models have advanced significantly in machine reading comprehension but often exhibit biases that hinder their performance, particularly with complex queries in adversarial conditions. This study evaluates the ELECTRA-small model on the Stanford Question Answering Dataset (SQuAD) v1.1 and adversarial datasets AddSent and AddOneSent. By identifying errors related to lexical bias, numerical reasoning, and entity recognition, we develop a multi-domain debiasing framework incorporating knowledge distillation, debiasing techniques, and domain expansion. Our results demonstrate up to 2.6 percentage point improvements in Exact Match (EM) and F1 scores across all test sets, with gains in adversarial contexts. These findings highlight the potential of targeted bias mitigation strategies to enhance the robustness and reliability of natural language understanding systems.

Key facts. No citing papers in the map yet (data through 7 Oct 2026). Builds on 15 of its 20 references that are in the map. First posted on arXiv 1 Jan 2026.

Builds on

15 of 20 references.

References · 15

  1. 2018BERT: Pre-training of Deep Bidirectional Transformers for Language Understandingcited by 40,204
  2. 2015Distilling the Knowledge in a Neural Networkcited by 10,076
  3. 2016SQuAD: 100,000+ Questions for Machine Comprehension of Textcited by 4,115
  4. 2018HotpotQA: A Dataset for Diverse, Explainable Multi-hop Question Answeringcited by 3,686
  5. 2017TriviaQA: A Large Scale Distantly Supervised Challenge Dataset for Reading Comprehensioncited by 2,895
  6. 2017Adversarial Examples for Evaluating Reading Comprehension Systemscited by 568
  7. 2019Don't Take the Easy Way Out: Ensemble Based Methods for Avoiding Known Dataset Biasescited by 279
  8. 2019MRQA 2019 Shared Task: Evaluating Generalization in Reading Comprehensioncited by 183
  9. 2020Self-Distillation Amplifies Regularization in Hilbert Spacecited by 171
  10. 2019End-to-End Bias Mitigation by Modelling Biases in Corporacited by 127
  11. 2019Unlearn Dataset Bias in Natural Language Inference by Fitting the Residualcited by 119
  12. 2020Mind the Trade-off: Debiasing NLU Models without Degrading the In-distribution Performancecited by 85
3 more
  1. 2021On the Efficacy of Adversarial Data Collection for Question Answering: Results from a Large-Scale Randomized Studycited by 30
  2. 2020Improving QA Generalization by Concurrent Modeling of Multiple Biasescited by 12
  3. 2024Adversarial Feature Alignment: Balancing Robustness and Accuracy in Deep Learning via Adversarial Trainingcited by 5

Built on by

Not cited yet.