# princeton-nlp/SimCSE

[EMNLP 2021] SimCSE: Simple Contrastive Learning of Sentence Embeddings https://arxiv.org/abs/2104.08821

Repository: https://github.com/princeton-nlp/SimCSE
Canonical: https://ross.abutalabs.com/products/simcse
Language: Python
License: MIT
License Family: permissive
Topics: nlp, sentence-embeddings
Last push: 2024-10-16T14:35:29+00:00

## Health v2 (maintenance only)
Score: 23/100 (v2, computed 2026-09-02T17:46:02.011165+00:00)
- activity 0, release rhythm 8, longevity 100
- inputs: {"age_days": 1965, "days_push": 686, "days_rel": null, "gap_med": null, "n_releases_24m": 0}
- flags: none
- formula: round(0.45*activity + 0.35*rhythm + 0.20*longevity); archived -> min(score, 10)

## Adoption (not part of the score)
Stars 3654, forks 536 (observed 2026-08-28T04:08:13.137192+00:00)

## What it is
SimCSE is a Python library and research codebase implementing simple contrastive learning for sentence embeddings, with pre-trained unsupervised and supervised models. It provides an easy-to-use sentence embedding tool compatible with Hugging Face Transformers.

## Use cases
- encode sentences into embeddings
- compute sentence similarity
- train custom sentence embedding models with contrastive learning
- evaluate sentence embeddings on STS benchmarks
- use pretrained sentence embeddings for semantic search

## When to choose
- you need high-quality sentence embeddings from a well-cited research model
- you want to train or fine-tune sentence embeddings with contrastive learning
- you want a simple pip-installable embedding tool

## When to avoid
- you need actively developed features or recent model architectures
- you need multilingual embeddings out of the box
- you need a production embedding service rather than a research library

## Facets
- artifact type: library
- maturity: maintenance
- function: nlp, machine-learning, deep-learning
- domain: machine-learning, deep-learning
- platform: python
- tags: sentence-embeddings, contrastive-learning, pretrained-models, research, natural-language-processing

## Member repositories
- princeton-nlp/SimCSE (main) score 23

## Provenance
- Observed fields: from GitHub, fetched 2026-08-28T04:08:13.137192+00:00.
- Health v2: computed from the inputs above; adoption is never an input.
- Inferred fields (summary, facets, guidance): AI-extracted, prompt v1, taxonomy v1, on 2026-08-29T18:31:39.897935+00:00, confidence not recorded.
  - readme: https://github.com/princeton-nlp/SimCSE (fetched 2026-08-28T04:08:13.137192+00:00, sha c3b64a9f579f)
  - registry_pypi: https://pypi.org/pypi/simcse/json (fetched 2026-08-29T09:25:49.568362+00:00, sha fc38eb8b6490)
- Data as of 2026-08-30T08:39:29.467469+00:00.