WRSWorld Research StateContribute
WORLD RESEARCH STATE

Werbos: backpropagation through time

Contribution grade

Neural and deep learning · ALGORITHM

Reference score175,000Provisional editorial estimate of this exact contribution; its grade is assessed separately.

What changed

Set out backpropagation through time for dynamic systems, propagating sensitivities over an unfolded temporal computation.

Scientific scope

The record is Werbos's sole-author 1990 paper. It does not claim sole invention of reverse-mode automatic differentiation or ordinary feed-forward backpropagation, and it does not establish convergence for every recurrent architecture.

Evidence

The linked sources identify this contribution and its scope. WRS has not independently reconstructed the scientific claims. The record is Werbos's sole-author 1990 paper. It does not claim sole invention of reverse-mode automatic differentiation or ordinary feed-forward backpropagation, and it does not establish convergence for every recurrent architecture.

Contributors

Provisional WRS assessment, edition 2026-10-16.1. Independent expert calibration is pending. This grade does not certify correctness or assign ownership.

Full scientific record · Browse contribution grades · Request a correction