Watch the next episode of #ResearchBytes, where @andreea_gane shares insights from work done in collaboration with @emblebi on ProtNLM, which can predict a short functional description from a protein’s amino acid sequence ↓
MACHINE LEARNING
-
Meta Launches DataPerf Platform for Data-Centric AI Leaderboards
By
–
Today, Meta researchers together with @MLCommons working group, are launching DataPerf, the first platform for building data & data-centric AI algorithm leaderboards. We're excited for how DataPerf will help to push the data-centric AI field forward
-
ICML2023 Sets Reviewer Expectations on Author Response Engagement
By
–
#ICML2023 draws a line in the sand: reviewers are expected to read author response, but aren't required to engage with them further. Very sensible position. Reviewing is a volunteer position. It shouldn't turn into a full-time job arguing the merits of a paper with the authors.
-
Researcher Shares 45 Papers in AI and Machine Learning
By
–
Numbers are just numbers but some stats are fun. Over the past 3.3 years, I co-authored ~45 papers , ~16 as lead author. Amongst my fav papers I led are UL2, U-PaLM, DSI, Synthesizer, Charformer and Long Range Arena. The most delicious paper I wrote was Flan .
-
New Method Grows Smaller Language Models Into Larger Ones
By
–
New method “grows” smaller language models that were already trained into larger ones. The technique performs better than similar methods, saving about 50% of the computational cost required to train a large model: https://
bit.ly/3JX3hw6 -
GPT-4 Shows Sparks of AGI: Expert Analysis
By
–
Join our LinkedIn Newsletter and explore the world of AI with us! Our Head of Machine Learning, Bram Verhoef, weighs in on @MSFTResearch #AI scientists' claim that GPT-4 demonstrates hints of AGI. Is it true? Find out more: https://
linkedin.com/pulse/gpt-4-sh
owing-sparks-agi-axelera-ai/
… #MachineLearning #AskAxeleraAI -
Barracuda Uses ML and AI to Detect Phishing Attacks
By
–
#Fact: 74% of organizations globally have fallen victim to phishing attacks With the help of using #ML on #Lakehouse, @Barracuda has built a powerful #AL engine using behavioral analysis to detect attacks and keep malicious actors at bay. See how
-
Time-Series Experimentation Production Workflow LinkedIn Live
By
–
We will be live on LinkedIn tomorrow at 12pm EST! Join Senior Data Scientist Andrii Kruchko to hear about the building blocks of a time-series experimentation and production workflow. #aiaccelerators https://
linkedin.com/events/7046903
399572783104/comments/
… -
ML papers should be trivial theory papers complicated
By
–
When writing an ML paper, it should be so trivial that all reviewers understand everything. When writing a theory paper, it should be so complicated that no reviewer understands everything. https://
x.com/shortstein/sta
/shortstein/status/1641111660248330241
…