mlsanityData to 14 Oct 2026

Multi-passage BERT: A Globally Normalized BERT Model for Open-domain Question Answering

arXiv 1908.08167 ↗First posted 22 Aug 2019cs.CL cs.AIcited by 96

Not on the map. Why?

Abstract (arXiv)

BERT model has been successfully applied to open-domain QA tasks. However, previous work trains BERT by viewing passages corresponding to the same question as independent training instances, which may cause incomparable scores for answers from different passages. To tackle this issue, we propose a multi-passage BERT model to globally normalize answer scores across all passages of the same question, and this change enables our QA model find better answers by utilizing more passages. In addition, we find that splitting articles into passages with the length of 100 words by sliding window improves performance by 4%. By leveraging a passage ranker to select high-quality passages, multi-passage BERT gains additional 2%. Experiments on four standard benchmarks showed that our multi-passage BERT outperforms all state-of-the-art models on all benchmarks. In particular, on the OpenSQuAD dataset, our model gains 21.4% EM and 21.5% F1 over all non-BERT models, and 5.8% EM and 6.5% F1 over BERT-based models.

Key facts. Cited by 96 papers in the map (data through 14 Oct 2026). First posted on arXiv 22 Aug 2019.

Builds on

No reference list yet.

Built on by

Cited by year (2026 to 14 Oct)

Citing papers per year: 2021: 27; 2022: 18; 2023: 21; 2024: 18; 2025: 9; 2026: 3.

96 papers.

Sort:

By area:

  1. 2022Atlas: Few-shot Learning with Retrieval Augmented Language Modelscited by 952
  2. 2022RealTime QA: What's the Answer Right Now?cited by 244
  3. 2021Efficient Training of Audio Transformers with Patchoutcited by 226
  4. 2021Retrieving and Reading: A Comprehensive Survey on Open-domain Question Answeringcited by 178
  5. 2021RocketQAv2: A Joint Training Method for Dense Passage Retrieval and Passage Re-rankingcited by 176
  6. 2023Evaluating Correctness and Faithfulness of Instruction-Following Models for Question Answeringcited by 151
  7. 2023How Can Recommender Systems Benefit from Large Language Models: A Surveycited by 142
  8. 2021End-to-End Training of Multi-Document Reader and Retriever for Open-Domain Question Answeringcited by 133
  9. 2024Transformers are Multi-State RNNscited by 120
  10. 2021TopiOCQA: Open-domain Conversational Question Answering with Topic Switchingcited by 96
  11. 2023Learning to Filter Context for Retrieval-Augmented Generationcited by 83
  12. 2022Efficient Long-Text Understanding with Short-Text Modelscited by 71
38 more
  1. 2021KG-FiD: Infusing Knowledge Graph in Fusion-in-Decoder for Open-Domain Question Answeringcited by 67
  2. 2021Natural Language Video Localization: A Revisit in Span-based Question Answering Frameworkcited by 55
  3. 2024Blinded by Generated Contexts: How Language Models Merge Generated and Retrieved Contexts When Knowledge Conflicts?cited by 48
  4. 2023A Survey on Long Text Modeling with Transformerscited by 47
  5. 2021Multi-task Retrieval for Knowledge-Intensive Taskscited by 44
  6. 2021UnitedQA: A Hybrid Approach for Open Domain Question Answeringcited by 42
  7. 2024VidMuse: A Simple Video-to-Music Generation Framework with Long-Short-Term Modelingcited by 40
  8. 2024TokenSelect: Efficient Long-Context Inference and Length Extrapolation for LLMs via Dynamic Token-Level KV Cache Selectioncited by 39
  9. 2025Parameters vs. Context: Fine-Grained Control of Knowledge Reliance in Language Modelscited by 39
  10. 2021R2-D2: A Modular Baseline for Open-Domain Question Answeringcited by 37
  11. 2022An Efficient Memory-Augmented Transformer for Knowledge-Intensive NLP Taskscited by 36
  12. 2024Improving Medical Reasoning through Retrieval and Self-Reflection with Retrieval-Augmented Large Language Modelscited by 34
  13. 2024Grounding Language Model with Chunking-Free In-Context Retrievalcited by 30
  14. 2021Adaptive Information Seeking for Open-Domain Question Answeringcited by 29
  15. 2024Selective Attention Improves Transformercited by 26
  16. 2021Dual Reader-Parser on Hybrid Textual and Tabular Evidence for Open Domain Question Answeringcited by 25
  17. 2024OLAPH: Improving Factuality in Biomedical Long-form Question Answeringcited by 20
  18. 2024Phaseformer: Phase-based Attention Mechanism for Underwater Image Restoration and Beyondcited by 15
  19. 2025TempRetriever: Fusion-based Temporal Dense Passage Retrieval for Time-Sensitive Questionscited by 15
  20. 2021Attention-guided Generative Models for Extractive Question Answeringcited by 14
  21. 2022A Thousand Words Are Worth More Than a Picture: Natural Language-Centric Outside-Knowledge Visual Question Answeringcited by 13
  22. 2023Normalization Enhances Generalization in Visual Reinforcement Learningcited by 13
  23. 2024Cross-modal Retrieval for Knowledge-based Visual Question Answeringcited by 13
  24. 2025Training a Utility-based Retriever Through Shared Context Attribution for Retrieval-Augmented Language Modelscited by 13
  25. 2021Contrastive Fine-tuning Improves Robustness for Neural Rankerscited by 12
  26. 2022From Easy to Hard: Two-stage Selector and Reader for Multi-hop Question Answeringcited by 12
  27. 2021Open-Retrieval Conversational Machine Readingcited by 11
  28. 2022Entity-aware Transformers for Entity Searchcited by 10
  29. 2023Legal Question-Answering in the Indian Context: Efficacy, Challenges, and Potential of Modern AI Modelscited by 10
  30. 2021More Than Reading Comprehension: A Survey on Datasets and Metrics of Textual Question Answeringcited by 8
  31. 2021Encoder Adaptation of Dense Passage Retrieval for Open-Domain Question Answeringcited by 8
  32. 2022Zemi: Learning Zero-Shot Semi-Parametric Language Models from Multiple Taskscited by 8
  33. 2022Metric-guided Distillation: Distilling Knowledge from the Metric to Ranker and Retriever for Generative Commonsense Reasoningcited by 8
  34. 2021Cheap and Good? Simple and Effective Data Augmentation for Low Resource Machine Readingcited by 7
  35. 2022Interactive Question Answering Systems: Literature Reviewcited by 7
  36. 2025Serialized EHR make for good text representationscited by 7
  37. 2021Pruning the Index Contents for Memory Efficient Open-Domain QAcited by 6
  38. 2021FLiText: A Faster and Lighter Semi-Supervised Text Classification with Convolution Networkscited by 5

50 most cited of 96 All 96 (JSON)