Results for "trigger-based behavior"

System-level behavior arising from interactions.

Research ensuring AI remains safe.

Reinforcement learning from human feedback: uses preference data to train a reward model and optimize the policy.

Artificial environment for training/testing agents.

Directly optimizing control policies.

Learning only from current policy’s data.

Acting to minimize surprise or free energy.

Internal representation of the agent itself.

Exact likelihood generative models using invertible transforms.

Strategy mapping states to actions.

Learning from data generated by a different policy.

Extracting system prompts or hidden instructions.

Interleaving reasoning and tool use.

Assigning a role or identity to the model.

Learning physical parameters from data.

Imagined future trajectories.

Inferring human goals from behavior.

Closed loop linking sensing and acting.

Fabrication of cases or statutes by LLMs.

Risk of incorrect financial models.

Rules governing auctions.

AI tacitly coordinating prices.

Early signals disproportionately influence outcomes.

Combines value estimation (critic) with policy learning (actor).

Learns the score (∇ log p(x)) for generative sampling.

Simple agent responding directly to inputs.

Dynamic resource allocation.

Continuous loop adjusting actions based on state feedback.

Algorithm computing control actions.

A mismatch between training and deployment data distributions that can degrade model performance.