RESEARCH - AI Dynamics

Clarification on longer-context models and future evaluation possibilities

By

–

11 July 2023 2h52

Yes, sorry I see what you mean. I didn’t mean to imply that they studied these longer-context models like LongNet above. I was just referencing that we talked about these models (it would have been impossible for the authors to eval it at that time, but hopefully in the future)

→ View original post on X — @rasbt

11 July 2023

TMLR Launches Infinite Conference Platform for Research Papers

By

@hugo_larochelle

–

11 July 2023 0h32

Please check out our latest TMLR innovation: the TMLR Infinite Conference! https://
tmlr.infinite-conf.org It's a site like the virtual sites of ML conferences, but that we continuously populate with TMLR papers that have a video.
Thanks @PaulVicol for the idea and website!

→ View original post on X — @hugo_larochelle

11 July 2023

Homomorphic Encryption Enhances Blockchain Privacy Solutions

By

@terence_mills

–

10 July 2023 23h00

How Homomorphic Encryption Can Make Blockchain Private
#blockchaintechnology #blockchains #blockchainrevolution @IIoT_World @ipfconline1 @jblefevre60 @JimMarous @insurtechtalk @KirkDBorne @kuriharan @LouisSerge http://
ow.ly/ca2o30sw8Ba

→ View original post on X — @terence_mills

10 July 2023

spaCy NLP: Game-Changer for Unstructured Data Processing

By

@dominodatalab

–

10 July 2023 22h47

Discover the power of #NLP with spaCy! According to Nikolay Manchev in @InfoWorld
, it's a game-changer for processing vast text volumes. Unleash the full potential of unstructured data today! https://
domino.buzz/44iOkNJ @NYIke #DataScience #AI

→ View original post on X — @dominodatalab

10 July 2023

Open-source collaboration to build competitive LLMs

By

@replit

–

10 July 2023 22h01

We need more collaboration in the open-source community to help produce LLMs that can go toe-to-toe with GPT-4.

To learn more watch the full AI panel with MosaicML’s Chief Scientist Jonathan Frankle in conversation with Replit’s Amjad Masad and Michele Catasta. pic.twitter.com/VeajA41d6d
— Replit ⠕ (@Replit) 10 juillet 2023

We need more collaboration in the open-source community to help produce LLMs that can go toe-to-toe with GPT-4. To learn more watch the full AI panel with MosaicML’s Chief Scientist Jonathan Frankle in conversation with Replit’s Amjad Masad and Michele Catasta.

→ View original post on X — @replit

10 July 2023

Quality Chat Datasets Essential for Open Source LLM Improvement

By

@replit

–

10 July 2023 22h01

We need good quality chat data sets, not just those scraped from GPT-4, to improve the standard of open source LLMs.

Watch the full AI panel at https://t.co/KWfZYbRuUN pic.twitter.com/Xpn9n4wHjL
— Replit ⠕ (@Replit) 10 juillet 2023

We need good quality chat data sets, not just those scraped from GPT-4, to improve the standard of open source LLMs. Watch the full AI panel at https://
youtu.be/B-szEQsQ9yI

→ View original post on X — @replit

10 July 2023

Causal Inference with Logistic Regression Tutorial

By

@miketamir

–

10 July 2023 21h01

Causal inference with logistic regression | A. Solomon Kurz https://
bit.ly/3o943zm #AI #MachineLearning #DeepLearning #LLMs #DataScience

→ View original post on X — @miketamir

10 July 2023

Transition from Individual Research to Large Team Collaboration at OpenAI

By

@_jasonwei

–

10 July 2023 20h30

Moving from Google Brain to OpenAI, one of the biggest changes for me was the shift from doing individual/small-group research to working on a team with several dozen people. Specifically, working on a bigger team has led me to think more about UX for researchers. Some examples:

→ View original post on X — @_jasonwei

10 July 2023