> ML_LITERATURE // KARPUKHIN-2020-DENSE-PASSAGE-RETRIEVAL-OPEN-DOMAIN-QA_v1.0
Dense Passage Retrieval for Open-Domain Question Answering (DPR)
Vladimir Karpukhin, Barlas Oğuz, Sewon Min, Patrick Lewis, Ledell Wu, Sergey Edunov, Danqi Chen, Wen-tau Yih · Conference on Empirical Methods in Natural Language Processing (EMNLP) (2020)
algorithm2020industry-standardthirdPartyReproduced
Principal Contribution
Proved that a dual-encoder architecture mapping queries and passages into dense 768-d vectors outperforms classical Lucene BM25 sparse lexical matching.
Operational Relevance
Directly guides architectural decisions, alignment strategy, and serving infrastructure for task-question-answering, task-feature-extraction.
Assumptions
- Empirical distribution regularity holds and target domain adheres to pretraining linguistic/visual support
Limitations
- Resource scaling, inference memory requirements, and alignment robustness vary with model size and hardware topology
Connected Algorithms, Architectures & Tools
Related Algorithms:
Related Architectures:
Implementing Libraries: