# PrincetonML/SIF

sentence embedding by Smooth Inverse Frequency weighting scheme

Repository: https://github.com/PrincetonML/SIF
Canonical: https://ross.abutalabs.com/products/sif
Language: Python
License: MIT
License Family: permissive
Last push: 2019-07-23T19:22:37+00:00

## Health v2 (maintenance only)
Score: 32/100 (v2, computed 2026-09-03T02:20:16.233290+00:00)
- activity 0, release rhythm 35, longevity 100
- inputs: {"age_days": 3582, "days_push": 2598, "days_rel": null, "gap_med": null, "n_releases_24m": 0}
- flags: no_releases
- formula: round(0.45*activity + 0.35*rhythm + 0.20*longevity); archived -> min(score, 10)

## Adoption (not part of the score)
Stars 1085, forks 303 (observed 2026-08-28T04:03:31.903973+00:00)

## What it is
A Python research library implementing the Smooth Inverse Frequency (SIF) weighting scheme for computing sentence embeddings, from the ICLR 2017 paper 'A Simple but Tough-to-Beat Baseline for Sentence Embeddings'. It includes scripts for generating embeddings, evaluating textual similarity, and training on supervised tasks.

## Use cases
- compute sentence embeddings with SIF weighting
- evaluate textual similarity between sentences
- reproduce results from the SIF paper
- build a simple baseline for sentence representation tasks
- train models on sentence similarity and entailment tasks

## When to choose
- you need a simple, strong baseline for sentence embeddings
- you want to reproduce or build on the SIF research
- you're working with GloVe word embeddings and need sentence-level vectors

## When to avoid
- you need modern transformer-based sentence embeddings like Sentence-BERT
- you want a production-ready, actively maintained library
- you need support for recent deep learning frameworks instead of Theano/Lasagne

## Facets
- artifact type: library
- maturity: maintenance
- function: nlp, machine-learning
- domain: machine-learning
- platform: python
- tags: sentence-embeddings, smooth-inverse-frequency, textual-similarity, research-code, word-embeddings, natural-language-processing, research

## Member repositories
- PrincetonML/SIF (main) score 32

## Provenance
- Observed fields: from GitHub, fetched 2026-08-28T04:03:31.903973+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 2026-08-30T06:49:41.661319+00:00, confidence not recorded.
  - readme: https://github.com/PrincetonML/SIF (fetched 2026-08-28T04:03:31.903973+00:00, sha 1547248ec4ce)
- Data as of 2026-08-30T08:39:29.467469+00:00.