Wikitopia
Dense Passage Retrieval
conceptNLP Technique
Try in Playground →
Overview
Developed byFacebook AI Research
Founded2020
LicenseCC-BY-NC 4.0
Open source✓ Open Source
Use caseretrieval-augmented question answering
Knowledge graph stats
Claims11
Avg confidence90%
Avg freshness99%
Last updatedUpdated 5 days ago
Trust distribution
100% unverified
Governance
Not assessed
Contribute governance data →

Dense Passage Retrieval

concept

Method for retrieving relevant passages using dense vector representations for question answering.

Compare with...

developed by

ValueTrustConfidenceFreshnessSources
Facebook AI Research○UnverifiedHighFresh1

primary use case

ValueTrustConfidenceFreshnessSources
retrieval-augmented question answering○UnverifiedHighFresh1

open source

ValueTrustConfidenceFreshnessSources
true○UnverifiedHighFresh1

founded year

ValueTrustConfidenceFreshnessSources
2020○UnverifiedHighFresh1

based on

ValueTrustConfidenceFreshnessSources
BERT encoder architecture○UnverifiedHighFresh1

alternative to

ValueTrustConfidenceFreshnessSources
BM25 sparse retrieval○UnverifiedHighFresh1

requires

ValueTrustConfidenceFreshnessSources
dense vector representations○UnverifiedModerateFresh1

license type

ValueTrustConfidenceFreshnessSources
CC-BY-NC 4.0○UnverifiedModerateFresh1

supports model

ValueTrustConfidenceFreshnessSources
bi-encoder architecture○UnverifiedModerateFresh1

integrates with

ValueTrustConfidenceFreshnessSources
FAISS vector search library○UnverifiedModerateFresh1

competes with

ValueTrustConfidenceFreshnessSources
TF-IDF retrieval methods○UnverifiedModerateFresh1

Alternatives & Similar Tools

BM25 sparse retrieval
alternative to
Compare →
TF-IDF retrieval methods
competes with
Compare →

Commonly Used With

FAISS vector search library

Related entities

Claim count: 11Last updated: 4/5/2026Edit history