# openai/generating-reviews-discovering-sentiment

Code for "Learning to Generate Reviews and Discovering Sentiment"

Repository: https://github.com/openai/generating-reviews-discovering-sentiment
Canonical: https://ross.abutalabs.com/products/generating-reviews-discovering-sentiment
Homepage: https://arxiv.org/abs/1704.01444
Language: Python
License: MIT
License Family: permissive
Topics: paper
Archived: true
Last push: 2023-06-28T23:45:25+00:00

## Health v2 (maintenance only)
Score: 10/100 (v2, computed 2026-09-03T02:20:16.233290+00:00)
- activity 0, release rhythm 35, longevity 100
- inputs: {"age_days": 3436, "days_push": 1162, "days_rel": null, "gap_med": null, "n_releases_24m": 0}
- flags: no_releases, archived
- formula: round(0.45*activity + 0.35*rhythm + 0.20*longevity); archived -> min(score, 10)

## Adoption (not part of the score)
Stars 1521, forks 375 (observed 2026-08-28T04:04:57.865989+00:00)

## What it is
Research code accompanying the OpenAI paper 'Learning to Generate Reviews and Discovering Sentiment', providing a byte-level multiplicative LSTM language model used as a text feature extractor. It includes a pretrained 4,096-unit model whose features contain a 'sentiment neuron' that achieves strong results on the Stanford Sentiment Treebank.

## Use cases
- extract features from text with a pretrained language model
- classify sentiment of product reviews
- reproduce the sentiment neuron paper results
- visualize the sentiment unit on the SST binary dataset
- experiment with unsupervised representation learning

## When to choose
- you want to reproduce or study the 2017 sentiment neuron research
- you need a character-level feature extractor for sentiment experiments
- you are exploring unsupervised disentangled representations

## When to avoid
- you need a maintained, production-ready NLP library
- you want modern transformer-based embeddings
- you need GPU support on current CUDA versions or recent Python compatibility

## Facets
- artifact type: library
- maturity: abandoned
- function: machine-learning, nlp, deep-learning
- domain: machine-learning, deep-learning
- platform: python
- tags: sentiment-analysis, language-model, feature-extraction, research-code, sentiment-neuron, archived, lstm, paper-code, natural-language-processing

## Member repositories
- openai/generating-reviews-discovering-sentiment (main) score 10

## Provenance
- Observed fields: from GitHub, fetched 2026-08-28T04:04:57.865989+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 2026-08-30T04:31:52.887718+00:00, confidence not recorded.
  - readme: https://github.com/openai/generating-reviews-discovering-sentiment (fetched 2026-08-28T04:04:57.865989+00:00, sha f4aa28d6da57)
  - homepage: https://arxiv.org/abs/1704.01444 (fetched 2026-08-29T11:34:57.588093+00:00, sha 7862b7a3d017)
  - site_page: https://info.arxiv.org/about/donate.html (fetched 2026-08-29T11:34:57.597079+00:00, sha cca9c3a11c56)
  - site_page: https://info.arxiv.org/about/ourmembers.html (fetched 2026-08-29T11:34:57.600707+00:00, sha 47cbc55ff1de)
  - site_page: https://info.arxiv.org/about (fetched 2026-08-29T11:34:57.602720+00:00, sha a1f16f915a9a)
  - site_page: https://info.arxiv.org/labs/index.html (fetched 2026-08-29T11:34:57.598961+00:00, sha b14a8d05a0ec)
- Data as of 2026-08-30T08:39:29.467469+00:00.