tensorflow / tensorflow/probability

Auto-diff for bijector jacobian

Open
#573 6 comments 0 reactions 0 assignees View on GitHub

Nobody has claimed this yet.

Dominant language
Jupyter Notebook
Stars
4.4k
Forks
1.1k
PR merge metrics
No merged PRs in 30d

Description

Forgive me if this is a dumb question because I'm new to this. Is implementing auto-diff for computing jacobian in the bijector superclass possible? Doing so should free the users from implementing the individual jacobian calculation in the subclasses once they have implemented the forward/inverse function.

Contributor guide

Open the contributing guide

First steps

  1. Read the whole issue, then the project's contributing guide.
  2. Comment on the issue to say you are picking it up — it saves two people doing the same work.
  3. Fork the repository and make your change on a branch.
  4. Open a pull request that references the issue number.

Research direction

No files or tests are named. Start by reading the bijector superclass and comparing subclass forward, inverse, and Jacobian implementations; determine whether automatic differentiation can cover the requested behavior. Done would be a decided, tested approach that removes redundant subclass Jacobian work, if feasible.

Written by the indexing model from the issue text.

Assessment

Tech stack
tensorflow
Domain
machine-learning
Issue type
Feature
Difficulty
5/5
Estimated time
Over a week
Activity status
Stale
Clarity
Needs clarification
Newbie friendliness
25/100

Get new issues in your inbox

A short digest of beginner-friendly GitHub issues.