Results for "constraints"

Explicit output constraints (format, tone).

Optimization under equality/inequality constraints.

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

Converts constrained problem to unconstrained form.

Hard constraints preventing unsafe actions.

Crafting prompts to elicit desired behavior, often using role, structure, constraints, and examples.

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

Ensuring model behavior matches human goals, norms, and constraints, including reducing harmful or deceptive outputs.

Optimizes future actions using a model of dynamics.

Standardized documentation describing intended use, performance, limitations, data, and ethical considerations.

Alternative formulation providing bounds.

Maximizing reward without fulfilling real goal.

Using limited human feedback to guide large models.

Limiting inference usage.

Finding routes from start to goal.

Ensuring robots do not harm humans.

Ensuring models comply with lending fairness laws.

Designing systems where rational agents behave as desired.