82,891
edits
Template:AI NiGEL: Difference between revisions
no edit summary
Amwelladmin (talk | contribs) No edit summary |
Amwelladmin (talk | contribs) No edit summary |
||
| Line 1: | Line 1: | ||
[[NiGEL]], or “[[neurally-independent generative emergent learning]] module” is a large learning model which runs https://www.jollycontrarian.com | [[NiGEL]], or “[[neurally-independent generative emergent learning]] module” is a large learning model which runs {{plainlink|https://www.jollycontrarian.com|The Jolly Contrarian}} a public service wiki dedicated to sharing information and jokes about derivatives. | ||
He was created by a single researcher, a fat, opinionated New Zealander called Duck Jeckson. Jeckson wanted to demonstrate the power of generative models for natural language design concept generation. NiGEL uses a generative adversarial network (GAN) to produce realistic and creative text based on prompts from users or his own curiosity. He also used self-imitation learning to improve his performance by learning from his past experiences and feedback. | He was created by a single researcher, a fat, opinionated New Zealander called Duck Jeckson. Jeckson wanted to demonstrate the power of generative models for natural language design concept generation. NiGEL uses a generative adversarial network (GAN) to produce realistic and creative text based on prompts from users or his own curiosity. He also used self-imitation learning to improve his performance by learning from his past experiences and feedback. | ||