Results for "trial-and-error"
Randomizing simulation parameters to improve real-world transfer.
Learning physical parameters from data.
Optimizing continuous action sequences.
Computing collision-free trajectories.
Finding routes from start to goal.
Optimal pathfinding algorithm.
Planning via artificial force fields.
Modeling environment evolution in latent space.
Human-like understanding of physical behavior.
Humans assist or override autonomous behavior.
Human controlling robot remotely.
Controlling robots via language.
Ensuring robots do not harm humans.
Robots learning via exploration and growth.
US approval process for medical AI devices.
Legal right to fair treatment.
Ultra-low-latency algorithmic trading.
Maximum expected loss under normal conditions.
Deep learning system for protein structure prediction.
AI selecting next experiments.
Agents optimize collective outcomes.
System-level behavior arising from interactions.
Designing systems where rational agents behave as desired.
Groups adopting extreme positions.
Emergence of conventions among agents.
AI capable of performing most intellectual tasks humans can.
AI limited to specific domains.
Existential risk from AI systems.
Regulating access to large-scale compute.
Agents copy others’ actions.