概念导读
Test-time scaling allocates more computation during inference through longer reasoning, search, sampling, verification, or multi-agent deliberation.
- Related
Test-Time Scaling
Test-time scaling allocates more computation during inference through longer reasoning, search, sampling, verification, or multi-agent deliberation.
For agents, test-time scaling often appears as workflow expansion or dynamic orchestration.
Related
- thinking-budget — User- or system-controlled reasoning budget.
- rl-conductor — Dynamic orchestration at inference time.
- mixture-of-agents — Ensemble-style test-time computation.