Model reference · open weights

scincl

Available as managed deployment Embeddings malteos · community Embeddings 1 variants 27k dl/mo

scincl is an open-weight embedding model from malteos. AxForge deploys and operates it for you on dedicated EU-owned hardware — with the licence handled where one is required.

Available as managed deployment — configured and operated for you on dedicated EU hardware, quoted per deployment.

What it is

Released bymalteos
TypeEmbedding models
TaskEmbeddings
Parameters (lead)110M
Context512 tokens
Runs withsentence-transformers
Released2022-03-02
Popularity27k downloads / month
LicenceOpen weights

About

What scincl is

SciNCL

SciNCL is a pre-trained BERT language model to generate document-level embeddings of research papers. It uses the citation graph neighborhood to generate samples for contrastive learning. Prior to the contrastive training, the model is initialized with weights from scibert-scivocab-uncased. The underlying citation embeddings are trained on the S2ORC citation graph.

Paper: Neighborhood Contrastive Learning for Scientific Document Representations with Citation Embeddings (EMNLP 2022 paper).

Read the full model card

Code: https://github.com/malteos/scincl

PubMedNCL: Working with biomedical papers? Try PubMedNCL.

How to use the pretrained model

Sentence Transformers

from sentence_transformers import SentenceTransformer

# Load the model
model = SentenceTransformer("malteos/scincl")

# Concatenate the title and abstract with the [SEP] token
papers = [
    "BERT [SEP] We introduce a new language representation model called BERT",
    "Attention is all you need [SEP] The dominant sequence transduction models are based on complex recurrent or convolutional neural networks",
]
# Inference
embeddings = model.encode(papers)

# Compute the (cosine) similarity between embeddings
similarity = model.similarity(embeddings[0], embeddings[1])
print(similarity.item())
# => 0.8440517783164978

Transformers

from transformers import AutoTokenizer, AutoModel

# load model and tokenizer
tokenizer = AutoTokenizer.from_pretrained('malteos/scincl')
model = AutoModel.from_pretrained('malteos/scincl')

papers = [{'title': 'BERT', 'abstract': 'We introduce a new language representation model called BERT'},
          {'title': 'Attention is all you need', 'abstract': ' The dominant sequence transduction models are based on complex recurrent or convolutional neural networks'}]

# concatenate title and abstract with [SEP] token
title_abs = [d['title'] + tokenizer.sep_token + (d.get('abstract') or '') for d in papers]

# preprocess the input
inputs = tokenizer(title_abs, padding=True, truncation=True, return_tensors="pt", max_length=512)

# inference
result = model(**inputs)

# take the first token ([CLS] token) in the batch as the embedding
embeddings = result.last_hidden_state[:, 0, :]

# calculate the similarity
embeddings = torch.nn.functional.normalize(embeddings, p=2, dim=1)
similarity = (embeddings[0] @ embeddings[1].T)
print(similarity.item())
# => 0.8440518379211426

Triplet Mining Parameters

SettingValue
seed4
triples_per_query5
easy_positives_count5
easy_positives_strategy5
easy_positives_k20-25
easy_negatives_count3
easy_negatives_strategyrandom_without_knn
hard_negatives_count2
hard_negatives_strategyknn
hard_negatives_k3998-4000

SciDocs Results

These model weights are the ones that yielded the best results on SciDocs (seed=4). In the paper we report the SciDocs results as mean over ten seeds.

modelmag-f1mesh-f1co-view-mapco-view-ndcgco-read-mapco-read-ndcgcite-mapcite-ndcgcocite-mapcocite-ndcgrecomm-ndcgrecomm-P@1Avg
Doc2Vec66.269.267.882.964.981.665.382.267.183.451.716.966.6
fasttext-sum78.184.176.587.975.387.474.688.177.889.652.51874.1
SGC76.882.777.28875.787.591.696.284.192.552.718.276.9
SciBERT79.780.750.773.147.771.148.371.749.772.652.117.959.6
SPECTER8286.483.691.584.592.488.394.988.194.853.92080
SciNCL (10 seeds)81.488.785.392.387.593.993.697.391.696.453.919.381.8
SciNCL (seed=4)81.289.085.392.287.794.093.697.491.796.554.319.681.9

Additional evaluations are available in the paper.

License

MIT

From the published model card. Full card on the HuggingFace links in the sidebar.

Using it via the API

Call it like any OpenAI endpoint

Once AxForge deploys scincl for you, it answers on the OpenAI-compatible API — the same base URL and keys as every other model. (scincl below is illustrative; you get the exact model name on deployment.)

$ curl -sS https://api.axforge.ai/v1/embeddings \
  -H "Authorization: Bearer $AXFORGE_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{"model":"scincl","input":"text to embed"}'

Create an account — your API key is available in the console. 3M free tokens every 30 days with every new account.

© 2026 AxForge · EU-hosted AI infrastructure Pricing Docs Trust Privacy Terms