(under construction)
These two papers introduce "attention" to NLP:
This paper is credited with introducing the "transformer":
GPT uses transformers to learn a language model.
Radford etal, Language Models are Unsupervised Multitask Learners, 2018.
This is a Guardian article written by GPT-3. A video by Tom Scott.
BERT is another transformer based language model:
More Applications of Transformers to NLP:
Transformers beat CNNs for image recognition:
Transformers for composing and performing music:
Proteinfolding:
Ethics:
Symbolic Regression with genetic algorithms (interpretable, bad at high-dimension problems) and deep learning (good at high-dimension problem). Data -> NN -> SR. [1] Afaiu, what is nice here is that the NN itself has a physical interpretation. See also AI Feynman.
AI Feynman.