Resilient backpropagation
Algorithms AIMech:ResilientBackpropagation DRAFT
Aliases: Rprop
A gradient-based optimization method that adapts coordinate step sizes using successive gradient signs and takes sign-directed parameter steps.
Identity Notes & Concept Scope
Rprop is an optimizer using backpropagated gradients, not the backpropagation procedure for computing those gradients. Update magnitude is controlled by step size rather than current gradient magnitude.
Claims & Grounded Evidence
1 Claim
Mechanism
#update-rule
Repeated gradient signs increase a coordinate's bounded step size, while a sign reversal decreases it; in the cited variant the reversal also suppresses that coordinate's current update.
⚠️ Curated Limitations & Negative Results
- Sign-reversal handling is scoped to the inspected PyTorch variant; other Rprop variants require comparison.
- No comparative speed or convergence claim is asserted.
Artificial Intelligence Ontology (AIO) Grounding
Mapped AIO Classes
| AIO Target Class | Mapping Relation | Mapping Rationale |
|---|---|---|
AIO:TrainingStrategy
TrainingStrategy
|
related_match | This algorithm can implement parameter fitting, but the pinned TrainingStrategy definition uses a preprocessing genus. This is a contextual training relationship, not equivalence or a subclass assertion. |
Evidence Sources & Primary Literature
1 Source| Source ID | Publication / Document | Type / Access | Reference & Checksum |
|---|---|---|---|
pytorch-docs
|
PyTorch contributors
|
official documentation
documentation
|
sha256:d6cf29f22549…
|
Curation Provenance
Curator: Codex
Method: ai_assisted
Model: GPT-6
Created: 2026-10-11T03:25:18Z
Updated: 2026-10-11T03:25:18Z
Source File:
data/curated/ResilientBackpropagation.yaml