Cadence

Website · Paper · PyPI · Documentation · Changelog

Research toward general intelligence through overlap consensus, equilibrium detuning and self-reflection.

Cadence's goal is a continuing learning system with the flexibility of animal and human problem solving: acquiring skills from experience, retaining useful knowledge, imagining alternatives and creating solutions across domains. The mission is to find the smallest persistent state and local update rule that can support these abilities. General intelligence is the research goal; the current library establishes bounded learning, memory and control results.

The organizing idea comes from Observer Patch Holography: bounded, observer-like patches with local state, ports, records, readback and repair. A disturbance exposes disagreement. The network can explore a possible response, test it against actual consequences and settle into a revised organization. We seek fewer mechanisms that solve more problems.

Three shared principles

• Overlap consensus: patches repair disagreement across their shared boundaries. The resulting equilibrium is an internally consistent model; its predictions still have to agree with experience.
• Equilibrium detuning: observed outcomes perturb that equilibrium. Local positive/negative contrasts change learned relationships; the same operation can adjust proposed actions while holding the model fixed.
• Functional self-reflection: patches can read internal state, predictions, proposed actions and unresolved mismatches through ordinary ports. Learning which internal summaries to read, how to feed them back and how to grow useful recursive organization is a central research direction.

The current implementation provides detached self-readback and private proposal revision. Automatically learned recursive hierarchies, curiosity and reliable creativity remain to be demonstrated. Creativity and self-reflection defines these goals and their behavioral tests.

Equilibrium world models

Cadence is being developed toward evolving equilibrium world models: brains whose internal representation of an actor in its world grows more accurate and expressive through experience. The representation should carry what is happening, what persists out of sight, what the actor controls and what its actions could cause. It need not describe those relationships in words to use them.

An equilibrium here need not mean motionless activity. A skilled actor can follow a coherent, changing trajectory of perceptions, expectations and actions. When events unfold as expected, the carried state should already be close to a useful interpretation of the next moment. Familiar danger can prompt a learned response immediately. Extra inference is needed when competing interpretations or consequential choices warrant it.

The proposed mechanism is recursive composition through bounded, observer-like, self-reading patches. Scene, body, candidate action and retrieved experience meet through learned nonlinear ports. One interpretation can become input to another, allowing the system to revise its understanding before acting. Actual evidence anchors that revision; changing the interpretation and learning new relationships are distinct operations. A random outcome can require a new response while remaining consistent with a correctly learned probability distribution. Teaching detuning uses a specified target to compute a learning signal; it is not synonymous with surprise.

Skilled demonstrations and the actor's own actions supply complementary experience. Watching reveals useful behavior and situations; acting tests what controls actually cause. The same learned relationships should support private branches that explore possible futures without altering factual memory. Plans must be judged by subsequent real outcomes. Grounded replay and sleep should then consolidate reliable relationships and successful decisions into cheaper habits, while unfamiliar situations can reopen deliberation.

This is the architectural vision. Current APIs provide components and bounded demonstrations, not the complete evolving world model. Each successive release is intended to take a concrete step toward this goal. Progress must be shown in prediction, retention, useful internal planning and behavior at declared resource budgets; a new version alone does not establish it. The equilibrium world-model guide explains the mechanism, the distinction between evidence repair and learning, and the current implementation boundaries.

Current library

Two primitives, composed through ports, and the belief patch that composes them toward a world model. The library requires Python 3.11+ and NumPy:

python -m pip install cadence-net

Start with the quickstarts: a record patch that learns a stream and sleeps, a settling brain that decides, and a temporal patch that learns a consequence and plans. The architecture guide maps each capability to its API and current scope; EquilibriumActor is a separate fixed linear-body component with exact Gaussian history compression.

Imagined continuations are isolated. Branches use the learned network without changing live activity, parameters or factual bookkeeping. Controlled experiments demonstrate useful planning. Creativity requires additional evidence that novel proposals satisfy meaningful constraints and survive actual evaluation; musical improvisation is one possible example.

Retained experience and new learning are tested together. Protected-path memory is conditional and finite. Importance is currently supplied; automatic relevance, selective forgetting, specialization and broad skill transfer remain research requirements. The task-design guide and common missteps explain how to measure them.

General mechanisms, different applications

Games, language, multimodal perception, embodied control and creative work should use the same learning and memory mechanisms with declared observation and action ports. The examples repository holds four worked applications: a worm that learns during its life, a composer that starts from silence, soft bodies that evolve together with their brains, and Connect Four. They are application tests, not definitions of the architecture. A result in one does not establish transfer to the others.

The scaling goal is better learned behavior from more experience and training, with as little manual design as possible. Measure unique experience, repeated training and model capacity separately while keeping port meanings and task evaluation fixed. The scaling guide defines these comparisons and the current computational limits.

Application demonstrations are published only when they establish their claimed behavior. Recall and interpolation are useful development tests; original creation requires stronger evidence. Every example states what is supplied, what is learned, what was measured and what it does not show, and carries a check that recomputes its numbers. Research receipts remain available with the paper without presenting those tests as finished products.

Proofs and compatibility

The bundled Lean library contains 169 checked conditional theorems about the mathematical components and their limits. It does not certify the complete Python implementation or prove intelligence. The paper identifies assumptions and reproducible evidence.

The PatchNet graph interface, GenericBrain, content memory, rehearsal and sequence readback are kept for the experiments that used them; they are distinct compositions, not parts of the two primitives. API reference.

Development installs use python -m pip install -e .. Optional backends apply to their documented graph APIs; the temporal implementation is NumPy. Pin a release or exact commit for reproducible work. MIT licensed.