Results for "policy"

Directly optimizing control policies.

Learning only from current policy’s data.

Learning from data generated by a different policy.

Optimizing policies directly via gradient ascent on expected reward.

Strategy mapping states to actions.

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

Continuous cycle of observation, reasoning, action, and feedback.

A learning paradigm where an agent interacts with an environment and learns to choose actions to maximize cumulative reward.

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

Stress-testing models for failures, vulnerabilities, policy violations, and harmful behaviors before release.

Rules and controls around generation (filters, validators, structured outputs) to reduce unsafe or invalid behavior.

Separates planning from execution in agent architectures.

Expected cumulative reward from a state or state-action pair.

Legal or policy requirement to explain AI decisions.

Algorithm computing control actions.

RL without explicit dynamics model.

Learning action mapping directly from demonstrations.

A high-priority instruction layer setting overarching behavior constraints for a chat model.

Model trained to predict human preferences (or utility) for candidate outputs; used in RLHF-style pipelines.

All possible configurations an agent may encounter.

Systematic review of model/data processes to ensure performance, fairness, security, and policy compliance.

Fundamental recursive relationship defining optimal value functions.

Formal framework for sequential decision-making under uncertainty.

Formal model linking causal mechanisms and variables.

Expected return of taking action in a state.

Expected causal effect of a treatment.

Ensuring learned behavior matches intended objective.

Randomizing simulation parameters to improve real-world transfer.

RL using learned or known environment models.

Learning policies from expert demonstrations.