Entity · technique

zero-shot learning

techniqueactivezero-shot-learning-50792bf9·1 events·first seen May 20, 2026

Aliases: zero-shot learning

Co-occurring entities

GPT-2 unsupervised language modeling OpenAI

More like this (12)

zero-shot evaluation few-shot learning One-Shot Imitation Learning zero-shot systematizer unsupervised learning Imitation Learning Unsupervised Pre-training Q-learning Language Models are Few-Shot Learners point cloud learning meta-learning soft-label learning

Recent events (1)

8Openai Blog·May 20, 2026·source ↗

Better language models and their implications

OpenAI announced GPT-2, a large-scale unsupervised language model capable of generating coherent multi-paragraph text and achieving state-of-the-art performance on language modeling benchmarks. The model demonstrated zero-shot capability across reading comprehension, machine translation, question answering, and summarization without task-specific fine-tuning. OpenAI notably withheld the full model release citing misuse concerns, marking an early high-profile instance of staged/responsible release policy.

Frontier Model Releases Evaluation and Benchmarking GPT-2 zero-shot learning unsupervised language modeling +3 more