skgrad

skgrad computes analytic input gradients for fitted scikit-learn models. It provides one consistent interface for affine estimators and multilayer perceptrons without numerical finite differences or automatic-differentiation frameworks.

import skgrad

values, jacobian = skgrad.value_and_jacobian(model, X)

values always has shape (samples, outputs) and jacobian has shape (samples, outputs, features). Regression values are predictions. Classification values are decision scores or logits, never probabilities.

Supported models

LinearRegression, Ridge, Lasso, and ElasticNet
LogisticRegression and RidgeClassifier
MLPRegressor with identity output
MLPClassifier, using pre-probability logits

Hidden MLP activations may be identity, logistic, tanh, or ReLU. ReLU uses a zero derivative at its nondifferentiable origin, matching scikit-learn's backpropagation convention.

API

skgrad.supports(model)
skgrad.model_output(model, X)
skgrad.input_jacobian(model, X)
skgrad.input_gradient(model, X, target=None)
skgrad.value_and_jacobian(model, X)

input_gradient is the scalar-output convenience API. A target is required when the model has multiple outputs. supports is the single capability check; users do not need to distinguish internal model families.

For downstream composition, skgrad.gradient_properties(model) reports computational metadata such as whether the input Jacobian is constant. This allows consumers to optimize integration without duplicating skgrad's estimator registry or exposing separate family-specific support predicates.

Tree models, parameter gradients, numerical differentiation, Integrated Gradients, and baseline handling are deliberately outside the package scope.

The value_and_jacobian primitive is designed for downstream composition. For example, UnifiedIG integrates its Jacobians to explain model outputs, and other consumers can combine them with analytic chain-rule transformations.

Download files

Download the file for your platform. If you're not sure which to choose, learn more about installing packages.

Source Distribution

skgrad-0.1.1.tar.gz (10.3 kB view details)

Uploaded Aug 3, 2026 Source

Built Distribution

If you're not sure about the file name format, learn more about wheel file names.

The dropdown lists show the available interpreters, ABIs, and platforms. Enable javascript to be able to filter the list of wheel files.

skgrad-0.1.1-py3-none-any.whl (7.5 kB view details)

Uploaded Aug 3, 2026 Python 3

File details

Details for the file skgrad-0.1.1.tar.gz.

File metadata

Download URL: skgrad-0.1.1.tar.gz
Upload date: Aug 3, 2026
Size: 10.3 kB
Tags: Source
Uploaded using Trusted Publishing? Yes
Uploaded via: twine/7.0.0 CPython/3.13.14

File hashes

Hashes for skgrad-0.1.1.tar.gz
Algorithm	Hash digest
SHA256	`da840971bf92493bcb9c8bbe836410b25ef30f7b5d93f7956db6df98e6c182db`
MD5	`ebfd567ed50deedcd88f174863f11b16`
BLAKE2b-256	`f9ca61352c59805f0ac69e43b3668cf48634f6f87f79ffdc16d4e528cf875199`

See more details on using hashes here.

Provenance

The following attestation bundles were made for skgrad-0.1.1.tar.gz:

Publisher: release.yml on LudgerHentschel/skgrad

Attestations: Values shown here reflect the state when the release was signed and may no longer be current.

Statement:
- Statement type: https://in-toto.io/Statement/v1
- Predicate type: https://docs.pypi.org/attestations/publish/v1
- Subject name: skgrad-0.1.1.tar.gz
- Subject digest: da840971bf92493bcb9c8bbe836410b25ef30f7b5d93f7956db6df98e6c182db
- Sigstore transparency entry: 2336315972
- Sigstore integration time: Aug 3, 2026
Source repository:
- Permalink: LudgerHentschel/skgrad@0eee0c8de1194ab3e11564c025b7a7178284178d
- Branch / Tag: refs/tags/v0.1.1
- Owner: https://github.com/LudgerHentschel
- Access: private
Publication detail:
- Token Issuer: https://token.actions.githubusercontent.com
- Runner Environment: github-hosted
- Publication workflow: release.yml@0eee0c8de1194ab3e11564c025b7a7178284178d
- Trigger Event: push