AI engineering glossary · LLM training alignment
What is test-time compute scaling?
Test-time compute scaling (also called Inference-time scaling): Improving answer quality by spending more computation at inference time, for example by generating longer reasoning, sampling many answers and voting, or searching with a verifier. It complements pretraining scale as a way to raise capability.
Test-time compute scaling: an example
Sampling 16 answers to a maths problem and picking the majority answer often beats a single answer from the same model.
Related terms
More LLM training alignment terms
- Trust region policy optimisation
- UltraFeedback
- Value function
- Warmup-stable-decay schedule
- Weak-to-strong generalisation
- ZeRO
- 3D parallelism
- 8-bit optimiser
Where you learn it
Test-time compute scaling is part of the AI Engineering Program, which teaches it with notes, a hands-on lab and an exam. The first lesson is free.