Back to events
Research paperCritical global significanceConfirmed confidence

GPT-3 demonstrates in-context learning at 175B parameters

OpenAI's 175B-parameter GPT-3 performs many tasks from instructions and a few examples in the prompt, without updating its weights.

Event details

Language Models are Few-Shot Learners described GPT-3, a 175-billion-parameter autoregressive Transformer trained at a scale that made zero-shot and few-shot prompting a practical interface. The model remained uneven and inherited major biases, but its ability to switch tasks through context helped establish the idea that a single large pretrained model could serve many applications without task-specific retraining.

Why it matters

GPT-3 made scale and in-context learning central to the modern foundation-model strategy and seeded a large API ecosystem.

98/100Global significance score. Regional effects are recorded only when the evidence supports a meaningful difference.