Models & architecture
Small language model
A relatively compact language model optimised for focused or efficient deployment.
Also known as: SLM
Definition
A small language model uses fewer parameters and generally lower compute than frontier LLMs, often targeting specific tasks or environments.
Why it matters
Smaller models can reduce cost, latency and data exposure, but capability must be evaluated on the exact banking task.
Related concepts
- Large language model
A foundation model trained to understand and generate sequences of language tokens.
- Model distillation
Training a smaller model to reproduce behaviours learned from a larger model.
- Quantisation
Representing model weights with lower numerical precision to reduce compute.