I gave a version of this talk "Scaling unlocks emergent abilities in language models" today at USC ISI. There is a video recording: https://
youtu.be/Z_Qt737HG-0 Thanks Justin Cho @HJCH0 for inviting me and organizing!
RESEARCH
-
Scaling Unlocks Emergent Abilities in Language Models Talk
By
–
-
Top 8 Machine Learning Algorithms to Study This Weekend
By
–
RT Here's a list of top 8 ML algorithms that you must go through over the weekend: http://
hubs.la/Q01xvDSf0
by @datasciencedojo #MachineLearning #Algorithms #DataScience -
Best AI Achievement 2023
By
–
This looks like the best thing in AI in 2023 so far👍 https://t.co/ActhWoiGhw
— Marek Rosa | European🇪🇺 | South African🇿🇦 (@marek_rosa) 3 février 2023This looks like the best thing in AI in 2023 so far
-
Topological Deep Learning Thesis Defense Recognition
By
–
Congratulations Chris. It was a magnificent thesis and defense. I recommend your thesis to everyone who wants to learn about topological deep learning!
-
Full Self Driving: Examining Five Years of Autonomous Vehicle Progress
By
–
Full Self Driving and the Emperor's New Clothes: @mpesce and co-hosts reflect on the promise and the progress over the past 5 years: https://
omny.fm/shows/the-next
-billion-seconds/the-next-billion-cars-autonomous-vehicles-fail?t=3m14s
… -
Google CEO announces LaMDA rollout, rival to ChatGPT
By
–
Google CEO @sundarpichai announced it is planning to roll out its LaMDA language model within "coming weeks and months” during live earnings call today. LaMDA = Language Model for Dialogue Applications. Google's version of ChatGPT.
-
LLM-Generated Scientific Contributions and Author Responsibility
By
–
At least within the context of the areas we work in, I don't really see the issue with that. If there are novel scientific contributions, which were created by the authors (potentially coerced from an LLM), and they are willing to take responsibility for it, seems fine to me.
-
6G Technology: Immersive Communication, Spatio-temporal Services, and AI
By
–
6G will significantly expand the technology landscape: we will see the emergence of immersive communication technologies, advanced Spatio-temporal services, and native compute-AI, among others. Source @ericsson Link > https://
bit.ly/3I31hRD via @antgrasso #6g #AI -
Flan-T5 Outperforms OPT-IML with Greater Efficiency
By
–
Check out OPT-IML, but a lot of people said they didn't know:
– Flan-T5 checkpoints are publicly available (without requesting access)
– Flan-T5 11B outperforms OPT-IML on MMLU and Big-Bench Hard, despite being 10x more compute efficient Checkpoints: https://
huggingface.co/docs/transform
ers/model_doc/flan-t5
… -
Dot Product and Cosine Similarity in NLP Explained
By
–
In this post, @luis_likes_math defines two types of similarities for sentences: dot product similarity and cosine similarity. These similarities are very useful in determining if two sentences are similar or different. Learn more about similarity in NLP in this exciting post: