# SeldonIO/alibi-detect

Algorithms for outlier, adversarial and drift detection

Repository: https://github.com/SeldonIO/alibi-detect
Canonical: https://ross.abutalabs.com/products/alibi-detect
Homepage: https://docs.seldon.io/projects/alibi-detect/en/stable/
Language: Jupyter Notebook
License: NOASSERTION
License Family: other
Topics: anomaly, outlier, concept-drift, detection, unsupervised-learning, semi-supervised-learning, adversarial, data-drift, drift-detection, time-series, images, tabular-data, text
Last push: 2025-12-11T16:35:15+00:00

## Health v2 (maintenance only)
Score: 55/100 (v2, computed 2026-09-02T17:46:02.011165+00:00)
- activity 56, release rhythm 28, longevity 100
- inputs: {"age_days": 2522, "days_push": 265, "days_rel": 265, "gap_med": null, "n_releases_24m": 1}
- flags: no_license
- formula: round(0.45*activity + 0.35*rhythm + 0.20*longevity); archived -> min(score, 10)

## Adoption (not part of the score)
Stars 2548, forks 253 (observed 2026-08-28T04:07:00.360987+00:00)

## What it is
Alibi Detect is a Python library providing algorithms for outlier, adversarial, and drift detection across tabular data, text, images, and time series. It supports both online and offline detectors with TensorFlow and PyTorch backends, aimed at monitoring machine learning models in production.

## Use cases
- detect data drift in production ML models
- find outliers in tabular datasets
- detect concept drift in streaming data
- spot adversarial attacks on image classifiers
- monitor distribution shifts in text data
- detect anomalies in time series
- set up drift detection for deployed models

## When to choose
- you need production-grade drift or outlier detection for ML pipelines
- you want both TensorFlow and PyTorch backend support
- you need detectors for mixed data types like images, text, and tabular data
- you are monitoring models in a MLOps setting

## When to avoid
- you need a general-purpose anomaly detection toolkit outside ML monitoring
- you require a permissive open-source license (it uses Business Source License 1.1)
- you need lightweight detectors without deep learning dependencies

## Facets
- artifact type: library
- maturity: active
- function: machine-learning, monitoring, analytics
- domain: machine-learning, data-science, artificial-intelligence, monitoring
- platform: python
- tags: outlier-detection, drift-detection, adversarial-detection, anomaly-detection, model-monitoring, tensorflow, pytorch, mlops

## Member repositories
- SeldonIO/alibi-detect (main) score 55

## Provenance
- Observed fields: from GitHub, fetched 2026-08-28T04:07:00.360987+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 2026-08-30T02:24:32.757295+00:00, confidence not recorded.
  - readme: https://github.com/SeldonIO/alibi-detect (fetched 2026-08-28T04:07:00.360987+00:00, sha f6cd986716c4)
  - registry_pypi: https://pypi.org/pypi/alibi-detect/json (fetched 2026-08-29T10:06:39.986359+00:00, sha 788388b636a6)
- Data as of 2026-08-30T08:39:29.467469+00:00.