Results for “Reasoning models”

1. glossary
   Reasoning model

   A model that spends additional compute at inference time to plan, verify and correct its answer before responding.

2. glossary
   Test-time compute

   Compute used when a model answers a query, as opposed to during training. Scaling it improves reasoning on hard problems.

3. glossary
   Sovereign AI

   National or regional efforts to control AI compute, models and data as strategic infrastructure.

4. glossary
   Distillation

   Training a smaller model to reproduce the behaviour of a larger one, transferring capability at lower cost.

5. glossary
   Inference

   Running a trained model to produce outputs. As agents operate continuously, inference cost becomes the dominant economic factor.

6. glossary
   Open-weight model

   A model whose trained parameters are published for download and use, with or without a fully open licence.

7. glossary
   World model

   A model that learns an internal representation of how an environment behaves, enabling prediction and planning.