Numenta Logo

FAQ

I’m new to HTM. Where can I learn more about it?

A great place to start is HTM School, an educational YouTube series. The series is designed to be viewed in order, with each episode explaining fundamental concepts. HTM School is aimed at a general audience. If you’re looking for more technical detail, you can view Numenta’s technical papers.

How can I join the HTM open source community?

There are many ways to get involved, but the first step is to join the HTM Forum, which is where you can introduce yourself, ask questions, find information and connect with others interested in and working on HTM. You can also join our Meetup Group and subscribe to our YouTube channel.

Where can I find the code?

You can find Numenta’s production and research code (and more) on Github. Community members often share their projects on HTM Forum or HTM Community Github.

What implementations are available?

We have multiple HTM implementations available. Numenta maintains and supports one in Python with C++. Community members have contributed implementations in languages such as Java, Clojure, Go, and JavaScript.

What are the licensing options?

Our open source code containing intellectual property is governed by an AGPL Version 3 license. For those who are unable to use AGPLv3, we offer a trial license, and for those interested in distributing the technology, we offer commercial licenses. You can read more about our licenses here.

Is there a specific application for HTM?

The algorithm lends itself well to high-speed temporal data, but it could potentially be applied to many different fields and endeavors. Have an idea? Discuss HTM applications on HTM Forum.

What is the current research focus for HTM?

Numenta’s current focus is developing a comprehensive theory of how the neocortex learns through movement — what is often called “sensorimotor” learning. You can view a recording of Numenta Open Source Community Manager Matt Taylor talking with Jeff Hawkins about this topic in this video series.