Cognition01
Systems that reason
Learning agents that plan, deliberate and know the edge of their own competence — trained to hold up under shift, not just on the benchmark.
- World models
- Reasoning & planning
- Calibrated uncertainty
Cognition · Machine
We build learning systems and the machines they command — one continuous stack from reasoning to real-world action.
01 — Mission
Founded as an independent laboratory, we publish openly, build the instruments we need, and hold our work to the standard of the physical result — not the demo.
The stack isn’t a metaphor. It’s a dependency graph — and every release runs the full length of it.
Learning agents that plan, deliberate and know the edge of their own competence — trained to hold up under shift, not just on the benchmark.
The software layer that turns a plan into dependable execution — agents that orchestrate tools, services and pipelines, with the control loops to recover when the real world pushes back.