Difficulty: Advanced

Combination of cooperation and competition.

Generator produces limited variety of outputs.

RL using learned or known environment models.

RL without explicit dynamics model.

Approximating expectations via random sampling.

Physical form contributes to computation.

Computing collision-free trajectories.

No agent benefits from unilateral deviation.

Controls amount of noise added at each diffusion step.

Measure of vector magnitude; used in regularization and optimization.

Emergence of conventions among agents.

Detecting and avoiding obstacles.

Control without feedback after execution begins.

Vectors with zero inner product; implies independence.

Intelligence and goals are independent.

Correctly specifying goals.

No agent can improve without hurting another.

Finding routes from start to goal.

Software pipeline converting raw sensor data into structured representations.

Software simulating physical laws.

Number of steps considered in planning.

Groups adopting extreme positions.

Directly optimizing control policies.

Updated belief after observing data.

Planning via artificial force fields.

Tendency to gain control/resources.

Belief before observing data.

Describes likelihoods of random variable outcomes.

Probability of treatment assignment given covariates.

Internal sensing of joint positions, velocities, and forces.