Artificial Intelligence #93

Hey, in this issue: recent advances in NLP via large pre-trained language models; scalable, low-cost training of massive deep learning models; a unified framework for model explanation; a dataset for fine-grained emotion classification; on the quality of auto-generated code; and more.

The sponsor of this issue is SambaNova.

• [ArXiv] Recent advances in NLP via large pre-trained language models: a survey
• [Project] Varuna: scalable, low-cost training of massive deep learning models (paper, code)
• [JMLR] Explaining by removing: a unified framework for model explanation
• Are we on the verge of chatting with whales?
• [Sponsored] On-demand webcast: Groundbreaking Solution for GPT

Learn how SambaNova has leveled the playing field, making large language models more accessible with the industry’s first subscription-based GPT solution for NLP. Watch this on-demand webcast to get an overview and demo of this groundbreaking new offering.

• [Data] GoEmotions: a dataset for fine-grained emotion classification
• Can I use this publicly available dataset to build commercial AI software? Most likely not
• A curated list of the latest breakthroughs in AI with a video explanation, link to the paper, and code
• On the quality of auto-generated code

The newsletter is now read by more than 520,000 subscribers. If you are building an AI or a data product or service, you are invited to join the list of?these great companies?by becoming a sponsor of one of the future newsletter issues. Feel free to reach out to?[email protected]?for more details on sponsorships.

Enjoy the newsletter? Help us make it bigger and better by sharing it with your colleagues and friends.

Have a nice weekend! See you next week. — Andriy