Index
·research·reading

Language Models are Few-Shot Learners

The GPT-3 paper that made in-context learning a central empirical phenomenon in modern AI.

paper · queued
Tom B. Brown et al.
arXiv:2005.14165
source ↗

This is the obvious bridge between scaling and interaction: a sufficiently large language model starts to absorb task specification from the prompt itself, turning examples into a temporary program rather than a conventional fine-tuning dataset.

Neighborhood

Related

Attention Is All You NeedAttention Is All You Ne...On the Measure of IntelligenceOn the Measure of Intel...Pretrained Transformers as Universal Computation EnginesPretrained Transformers...AI systems engineeringAI systems engineeringFull Stack Artificial IntelligenceFull Stack Artificial I...Reaching for the IntangibleReaching for the Intang...What is intelligence?What is intelligence?The APIThe APIFull-Stack Artificial IntelligenceFull-Stack Artificial Intel...Language Models are Few-Shot Le...