This English page is a curated translation layer linked to the Korean source article. Community changes are currently made on the Korean source, where the full revision history and anonymous edit trail are preserved.
GPT stands for Generative Pre-trained Transformer. The family popularized a workflow in which a Transformer language model is pre-trained on broad text data and then adapted for downstream tasks or interaction patterns.[1]
How it works
GPT models generate text by predicting the next token from prior context. Later generations expanded scale and in-context learning, allowing a single model to perform many tasks from instructions and examples without task-specific parameter updates.[2]
Why it matters
GPT is one of the most influential model families in the modern LLM ecosystem and forms part of the technical background of products such as ChatGPT. The name refers to a model family rather than a single permanent capability set.
Related concepts
SOURCES
Sources
KNOWLEDGE LINKS
Continue from here
BACKLINKS