Coming out of stealth · October 28Out of stealth · October 28
Cognitive infrastructure

The world's biggest problems start with a dissonance.

Cognitive infrastructure for continual reasoning under uncertainty.

Scroll
Who we are

Post-LLM research lab. Modeling thought. Building stateful, common-sense reasoning layers. Belief states that hold what organizations believe right now, with confidence, provenance and history, incentivized to find coherence and dissonance.

Origins
What we do
Form beliefsBeliefs with confidence and provenance.
Find dissonanceDissonance and contradiction detection.
Condition agencyHumans and agents interact with belief states.
Expertise
continual reasoning under uncertaintyinner world modelsreal-world representationsagent world modelsbelief revisionenergy based modelstheory of mindquantum physics inspired
counterfactual reasoningcausal inferencenash equilibriumgame theoreticuncertainty quantificationbayesian inferenceneurosciencereinforcement learningmachine learningmathematics

Developer
APIs

Endpoints for sessions, extraction, beliefs, dissonance and counterfactual reasoning.

Learn more →

Conversations
people trust

Belief conditioned reasoning and replies. See what the system believes, what changed, and where the evidence disagrees with itself.

Learn more →

Belief conditioned
agency

Agents hold belief states instead of system prompts, static config or memory files. Each department reasons. The company state is shared. Dyssonance decides who moves work forward and where the organization disagrees with itself.

Learn more →

Observability
and control

Watch the system change its mind. Clusters, coherence and dissonance edges, confidence and surprise per node.

Learn more →

Continual Reasoning
Under Uncertainty.