OpenAI logo

How OpenAI Uses Fine-tuning

6 engineering articles about Fine-tuning from OpenAI's engineering team

Articles

Filter:
OpenAI logo
OpenAI
Advanced
The article discusses emergent misalignment in large language models, particularly focusing on how misaligned persona features can lead to generalized misalignment.
OpenAI Team
16 min read
Has Summary
--
OpenAI logo
OpenAI
Intermediate
The article discusses the development of a neural network capable of playing Minecraft through a method called Video PreTraining (VPT), leveraging a large dataset of unlabeled gameplay videos.
Bowen Baker
8 min read
Has Summary
--
OpenAI logo
OpenAI
Advanced
The article discusses the lessons learned from deploying language models, focusing on safety and misuse.
Miles Brundage
14 min read
Has Summary
--
OpenAI logo
OpenAI
Advanced
The article discusses the application of reinforcement learning from human feedback to enhance the summarization capabilities of language models.
Nisan Stiennon
16 min read
Has Summary
--
OpenAI logo
OpenAI
Advanced
The article discusses the fine-tuning of the 774M parameter GPT-2 language model using human feedback to improve performance on various natural language tasks, including summarization and stylistic...
Daniel Ziegler
9 min read
Has Summary
--
OpenAI logo
OpenAI
Advanced
The article discusses the advancements in language models, particularly focusing on GPT-2, which generates coherent text and performs various language tasks without task-specific training.
Alec Radford
12 min read
Has Summary
--

You've reached the end! All 6 articles loaded.