gpt3-few-shot-2020

IN premise — entries/2026/06/21/wiki-Large_language_model-chunk-5.md

Created 2026-06-21T09:50:09+00:00

GPT-3 (Brown et al. 2020) demonstrated that few-shot learning emerges from scale, using 175B parameters without task-specific fine-tuning

Summary

GPT-3 showed that simply making a neural network large enough lets it pick up new tasks from a handful of examples written directly in the prompt, without any extra training step. This matters because it means general-purpose capability can emerge from raw scale alone, reducing the need to build and maintain separate specialized models for each task.

Dependents

These beliefs depend on this one:

IN gpt-series-demonstrated-exponential-capability-emergence — The GPT series demonstrated exponential capability emergence across four generations: basic language modeling (GPT-1, 117M params, 2018) → zero-shot multitask (GPT-2, 1.5B, 2019) → few-shot in-context learning (GPT-3, 175B, 2020) → multimodal reasoning (GPT-4, 2023).
OUT in-context-learning-reliably-replaces-finetuning — In-context learning reliably replaces per-task fine-tuning for most downstream applications, as demonstrated by GPT-3's few-shot capabilities operating without gradient updates.