Article-encoder role
According to the model card, this is the MedCPT Article Encoder, paired with a separate query encoder for query-to-article search.
Open Source Model Profile · ncbi
MedCPT-Article-Encoder
MedCPT-Article-Encoder is a 109M-parameter BERT biomedical encoder from ncbi. Its model card documents PubMed article embeddings for semantic search.
- Publisher
- ncbi
- Task
- feature-extraction
- Model type
- bert
- License
- public-domain
- Library
- transformers
- Publication status
- Accepted · not indexed
Model overview
MedCPT-Article-Encoder is published by ncbi as a feature-extraction model. The captured configuration identifies BertModel with model type bert, and Safetensors metadata reports 109,482,240 parameters. According to the model card, this repository contains the MedCPT Article Encoder for article embeddings, with other recorded as the license.
Recorded capabilities
Biomedical retrieval purpose
The model card describes embeddings of biomedical texts for semantic search and dense retrieval over PubMed-style content.
Reported pre-training scale
According to the model card, MedCPT was pre-trained on 255M query-article pairs from PubMed search logs, with strong zero-shot biomedical retrieval reported.
BERT 109M configuration
Captured config identifies BertModel and bert, with Safetensors metadata reporting 109,482,240 parameters.
Use cases in the source record
- Biomedical semantic search and dense retrieval over article titles and abstracts.
- According to the model card, query-to-article search when paired with the separate MedCPT Query Encoder.
Limitations and unknowns
- No independent evaluation results, context-window value, or hardware requirement was extracted.
- Scale and performance statements come from the publisher model card and were not independently verified by Ethen.
- Provider state is historical snapshot data, not independently refreshed current availability.
Source and provenance
Source: ncbi/MedCPT-Article-Encoder
Captured: Unknown. Processed: 2026-09-07T19:34:53.154843+00:00.
MedCPT Introduction MedCPT generates embeddings of biomedical texts that can be used for semantic search (dense retrieval) . The model contains two encoders: MedCPT Query Encoder : compute the embeddings of short texts (e.g., questions, search queries, sentences). MedCPT Article Encoder : compute the embeddings of articles (e.g., PubMed titles & abstracts). This repo contains the MedCPT Article Encoder. MedCPT has been pre-trained by an unprecedented scale of 255M query-article pairs from PubMed search logs , and has been shown to achieve state-of-the-art performance on several zero-shot biomedical IR datasets. In general, there are…
F001F002F003F004F005F006F007F008F010F011F012F013F014