Cognition · Data artifact
Search Budget Policy
Data artifactCognitionCognition & Memoryarc:SearchBudgetPolicy
A configuration artifact defining when a search planner must stop and return an action: fixed iteration count, wall-clock deadline, convergence criterion or a combination.
Responsibility. Bounds planning computation per decision.
Also known as: Computational budget, Simulation budget
Relationships
constrains control
Design guidance
- SHOULD combine a minimum iteration count with a time budget and convergence-based early stopping in production.
Quantitative guidance
As stated by the sources; verify before use.
- Examples: 10,000 simulations per action selection; 1000 ms deadline; stop early when the top action holds ~95% visit share (Ch5.5).
Classification
- Patterns
- Fixed iteration budgetTime-based cutoffAdaptive convergence-based early terminationMinimum iterations plus deadline or convergence
- Risks mitigated
- Shallow trees missing long-term consequencesStale real-time decisionsWasted computation on already-confident decisions
Sources
- Ch5.5: T. Nguyen, "Monte Carlo Tree Search Fundamentals," in Mastering Agentic AI Systems: Guide for the NVIDIA NCP-AAI Exam, 1st ed. 2026, ch. 5.5. ISBN: 9798244538229.