Eureka: Human-Level Reward Design via Coding Large Language Models - NASA/ADS

Retrieved on: 2024-01-29 19:58:09

Tags for this article:

Artificial intelligence

Deep learning

Large language models

OpenAI

Natural language processing

Reinforcement learning from human feedback

GPT-4

Generative pre-trained transformer

Reinforcement learning

Prompt engineering

Reinforcement

Click the tags to see associated articles and topics

Eureka: Human-Level Reward Design via Coding Large Language Models - NASA/ADS. View article details on HISWAI: https://ui.adsabs.harvard.edu/abs/2023arXiv231012931M/abstract

Excerpt

Large Language Models (LLMs) have excelled as high-level semantic planners for sequential decision-making tasks. However, harnessing them to learn ...

Article found on: ui.adsabs.harvard.edu

View Original Article