BLURB

BLURB evaluates biomedical language understanding across NER, PICO, relation extraction, sentence similarity, classification, and QA tasks.

21rows
blurb_score_macro_avgprimary metric
2026-05-06sampled

Metadata

ID
blurb
Category
Healthcare
Release
2020-07-31

Metrics

BLURB Score (Macro Avg.), Micro Avg., Named Entity Recognition, PICO, Relation Extraction, Sentence Similarity, Document Classification, Question Answering

Latest Results

Rows are ranked by BLURB macro-average score. Source model display names and organization labels are preserved.

Rank Subject BLURB Score (Macro Avg.) Model Match Provenance Sampled
1 BioLinkBERT-Large 84.30 — Imported 2026-05-06
2 BioM-ALBERT-xxlarge-PMC 84.10 — Imported 2026-05-06
3 BioM-ELECTRA-Large 83.81 — Imported 2026-05-06
4 BioM-BERT-PMC-Large 83.63 — Imported 2026-05-06
5 BioLinkBERT-Base 83.39 — Imported 2026-05-06
6 BioM-ALBERT-xxlarge 82.97 — Imported 2026-05-06
7 MSR BiomedBERT-LARGE (fine-tuning stabilization; uncased; abstracts) 82.91 — Imported 2026-05-06
8 MSR BiomedBERT (fine-tuning stabilization; uncased; abstracts) 82.75 — Imported 2026-05-06
9 BioELECTRA-Base 82.60 — Imported 2026-05-06
10 BioM-ELECTRA-Base 82.46 — Imported 2026-05-06
11 MSR BiomedBERT (uncased; abstracts + full text) 81.50 — Imported 2026-05-06
12 MSR BiomedBERT (uncased; abstracts) 81.16 — Imported 2026-05-06
13 BioBERT (cased) 80.34 — Imported 2026-05-06
14 Scibert (uncased) 78.86 — Imported 2026-05-06
15 Scibert (cased) 78.14 — Imported 2026-05-06
16 ClinicalBERT (cased) 77.29 — Imported 2026-05-06
17 RoBERTa (cased) 76.46 — Imported 2026-05-06
18 BlueBERT (cased) 76.27 — Imported 2026-05-06
19 BERT base (uncased) 76.11 — Imported 2026-05-06
20 BERT base (cased) 75.86 — Imported 2026-05-06
21 Unicoder base (multilingual) 73.60 — Imported 2026-05-06