Zubnet AILearnWiki › Annotation
Training

Annotation

Data Labeling, Data Annotation
The process of adding labels, tags, or metadata to raw data so it can be used for supervised learning. Annotating images means drawing bounding boxes around objects. Annotating text means labeling entities, sentiment, or intent. Annotating for RLHF means ranking model responses by quality. Annotation is the human labor that turns raw data into training data.

Why it matters

Annotation is the unglamorous foundation of supervised AI. Every labeled dataset, every fine-tuned model, every aligned assistant depends on human annotators who spent hours labeling data correctly. The quality of annotations directly determines model quality — inconsistent or biased labeling produces inconsistent and biased models. It's the most labor-intensive and often most expensive part of building AI systems.

Deep Dive

Annotation workflows typically involve: (1) creating clear labeling guidelines (what counts as "positive sentiment"? what's the boundary of a "car" in a bounding box?), (2) training annotators on the guidelines, (3) annotating data with multiple annotators per example (for quality control), (4) measuring inter-annotator agreement (do annotators agree on labels?), and (5) resolving disagreements (through adjudication or majority vote). Low agreement often indicates ambiguous guidelines or genuinely ambiguous data.

RLHF Annotation

For LLM alignment, annotation means comparing model responses: "Is response A or response B better for this prompt?" This preference annotation is particularly challenging because "better" is subjective, context-dependent, and culturally variable. Annotator demographics, expertise, and instructions all influence the resulting preference data, which in turn shapes model behavior. This is why alignment is often described as encoding the values of whoever writes the annotation guidelines.

AI-Assisted Annotation

Increasingly, AI models assist with annotation: pre-labeling data that humans then correct (faster than labeling from scratch), generating synthetic annotation data, or serving as additional annotators alongside humans. This creates an interesting feedback loop: AI helps label data that trains better AI. The risk is that AI-assisted annotation inherits the biases of the assisting model, so human oversight remains essential.

Related Concepts

← All Terms
← Alignment Anthropic →