Generative pre-trained transformer: Difference between revisions

From ACT Wiki
Jump to navigationJump to search
imported>Doug Williamson
(Create page - sources - Wikipedia https://en.wikipedia.org/wiki/Generative_pre-trained_transformer and ACT - https://www.treasurers.org/hub/treasurer-magazine/chatgpt-coming-treasurers)
 
(Add link.)
 
(5 intermediate revisions by 2 users not shown)
Line 8: Line 8:




This pre-training may then be supplemented by additional fine-tuning human supervised training, known as Reinforcement Learning from Human Feedback (RLHF).
GPT's unsupervised pre-training may then often be supplemented by additional fine-tuning human supervised training, known as Reinforcement Learning from Human Feedback (RLHF).




Line 17: Line 17:
* [[ChatGPT]]
* [[ChatGPT]]
* [[Enterprise-wide resource planning system]]
* [[Enterprise-wide resource planning system]]
* [[Generative AI]]  (GenAI)
* [[Google Gemini]]
* [[GPT-4]]
* [[Information technology]]
* [[Information technology]]
* [[Large language model]]  (LLM)
* [[Natural language]]
* [[Natural language]]
* [[Natural language processing]]
* [[Natural language processing]]
Line 25: Line 29:
* [[Software robot]]
* [[Software robot]]


[[Category:The_business_context]]
 
==Other resource==
*[https://cdn.openai.com/research-covers/language-unsupervised/language_understanding_paper.pdf Improving Language Understanding by Generative Pre-Training, Radford, Narasimhan, Salimans & Sutskever, 2018]
 
[[Category:Identify_and_assess_risks]]
[[Category:Identify_and_assess_risks]]
[[Category:Manage_risks]]
[[Category:Manage_risks]]
[[Category:Risk_reporting]]
[[Category:Risk_frameworks]]
[[Category:Risk_frameworks]]
[[Category:Risk_reporting]]
[[Category:The_business_context]]
[[Category:Technology]]

Latest revision as of 22:20, 11 May 2024

Information technology - software - natural language processing - artificial intelligence - chatbots.

(GPT).

Generative pre-trained transformers are language models that have been pre-trained on large datasets of unlabelled natural language text.

They can generate new text that is human-like, and in some cases may be difficult to distinguish from human-written text.


GPT's unsupervised pre-training may then often be supplemented by additional fine-tuning human supervised training, known as Reinforcement Learning from Human Feedback (RLHF).


See also


Other resource