Visit the lab · Models · Repositories · Community
Other ways to support: Cash App · PayPal · Venmo
TK Integers report · TAARDIS runtime
I'm a hands-on researcher in Alabama, building Internectual Labs from a laptop at home. I learn by experimenting, connect new ideas with published research, and use AI to help turn those ideas into working tools.
My focus is making AI more efficient while checking what happens to reasoning, memory use, and speed. I also build practical tools for Linux, Android, and the maker community.
Project details and technical links
Ternary computing & model conversion
Base-3 weight packing, selectable weight and scale formats, GPU kernel experiments, and an integrated Model Atlas.
Explore the project → · Read the reportTell your AI to get a GRIP.
Project continuation for OpenCode, with saved progress and a record of what changed, what was tested, and what still needs work.
Explore the project →Full-ternary model project
Published model files and documentation for the lab's full-ternary research. See the model card for usage and limitations.
View the model → · Runtime forkAndroid meets the maker bench
Install and monitor KlipperScreen on Android, with USB ADB forwarding and Wi-Fi fallback.
Explore the project →Sponsorship helps support GPU time, broader model evaluation, and continued development of the lab's public tools. If this work is useful to you, you can support it through GitHub Sponsors.
I'm welcoming collaborators interested in model evaluation, training pipelines, software tooling, and interface design. Start with a public project: reproduce a result, report a bug, improve the documentation, or propose a focused contribution.
Visit the lab · Browse my repositories · Join the community
Today's minds are always on the internet, but it's what you do with it that sets you apart.
— Cody Dixon



