representation

Definition

The internal vector encoding a model produces for an input — an embedding whose geometry is meant to reflect semantics, so similar inputs land near each other. Good representations transfer: downstream tasks can train small heads on top with far less data than end-to-end training would need. Representation learning studies how to obtain them, often without labels.

Explainers using this term

• Paper Walkthrough: The Design Fundamentals of Pixel Text Representation LearningOn the Design Fundamentals of Pixel Text Representation Learning
• Paper Explained: Beyond Data Scaling — Why the Backbone, Not the Trajectory Count, Decides Your VLA (VLAct)Beyond Data Scaling: Representation-Centric Continued Pre-training for Vision-Language-Action Models
• Mutual Information — Putting a Number on What You KnowRepresentation Learning with Contrastive Predictive Coding