Perplexity is a model evaluation metric for language models that measures how well a model predicts text; lower perplexity means less surprise…
Pose estimation in computer vision locates the keypoints and orientation of people or objects in images and video to describe their posture…
Precision in model evaluation is the share of a model's positive predictions that are actually correct, measuring how trustworthy its positive alerts…
Pretraining is the first, largest stage of model training, in which a neural network learns general patterns from vast unlabeled data before…
A prompt is the input text, instructions and context given to a language model, which the model continues or answers to produce…
Prompt caching reuses the computed attention state of a repeated prompt prefix, cutting latency and input cost when serving large language models.
Prompt injection is an attack in which crafted text makes a language model follow an attacker's instructions instead of those set by…