Inference
Inferenz
Inference refers to the process of applying an already-trained model to new input in order to produce an output — as opposed to training, where the model itself is adjusted. Every answer an AI gives in live operation is the result of an inference pass.
Source: NVIDIA — AI inference explained