Results for "instruction design"

Early architecture using learned gates for skip connections.

A narrow hidden layer forcing compact representations.

Using same parameters across different parts of a model.

Capabilities that appear only beyond certain model sizes.

Empirical laws linking model size, data, compute to performance.

All possible configurations an agent may encounter.

Set of all actions available to the agent.

Balancing learning new behaviors vs exploiting known rewards.

Separates planning from execution in agent architectures.

System that independently pursues goals over time.

Vectors with zero inner product; implies independence.

Converts constrained problem to unconstrained form.

Ensuring AI systems pursue intended human goals.

Correctly specifying goals.

Required descriptions of model behavior and limits.

Privacy risk analysis under GDPR-like laws.

Mechanism to disable AI system.

Startup latency for services.

Field combining mechanics, control, perception, and AI to build autonomous machines.

The physical system being controlled.

Algorithm computing control actions.

Study of motion without considering forces.

Robots learning via exploration and growth.

Systems where failure causes physical harm.

Mathematical guarantees of system behavior.

Predicting protein 3D structure from sequence.

AI discovering new compounds/materials.

Modeling chemical systems computationally.

Rules governing auctions.

Collective behavior without central control.