FAQ
Frequently asked questions about SignalWeave Modern
This page answers the practical questions new users usually ask before they decide whether to use the browser build, the desktop build, or the sample project.
What is SignalWeave Modern used for?
SignalWeave Modern is a cross-platform neural network trainer and analysis workbench. It is designed for people who want an application with project files, graphs, testing, pattern inspection, and analysis rather than just a code library.
What kinds of networks does it support?
Modern supports feed-forward networks with direct, 3-layer, and 4-layer topologies as well as simple recurrent networks.
What is stored in the project file?
The project file stores the network definition, embedded patterns, current weights, completed cycles, training sessions, rollback checkpoints, and related workspace state. Modern intentionally avoids separate network, pattern, and weight files in the main workflow.
Can I use SignalWeave in the browser?
Yes. The WebAssembly build is available at /app/. The desktop and browser editions share the same .NET engine and general workflow.
What does Train #2 or Train #3 mean?
It means continue training from the current weights for one more training block. It does not mean restart from scratch and it does not switch to a different algorithm.
Why are there training sessions and rollback?
Each explicit training run creates a visible checkpoint. That lets you compare weight states, inspect historical error graphs, and roll back to an earlier checkpoint if a later run performs worse.
What is the difference between Modern and Classic?
Modern is organized around one project file and integrated tabs for network settings, training, tests, weights, patterns, and analysis. Classic preserves the BasicProp-style workflow and is useful mainly if you want that exact reference surface.
Is SignalWeave a replacement for BasicProp?
Yes. SignalWeave was built as a maintained, cross-platform, open-source alternative after BasicProp was retired. Modern is the recommended workflow, while Classic preserves the reference-oriented BasicProp-style surface.
How do I start quickly?
Open the sample project, run Train #1, then inspect the result in Tests, Weights, and Analysis. The Quick Start guide walks through that sequence.
Is SignalWeave open source?
Yes. Source, releases, and project history are available on GitHub.