DARPA Research on Cryptocurrency and National Security Risks https://
oodaloop.com/archive/2022/1
1/08/darpa-research-on-cryptocurrency-and-national-security-risks/
… via @ooda
RESEARCH
-
DARPA Research on Cryptocurrency National Security Risks
By
–
-
Understanding Opportunity Cost: Swedish Economists’ Key Insight
By
–
A survey of every Swedish economist found the economics concept that would benefit the most people to understand is "opportunity cost." It is what you give up when you make a choice. Not just what you don’t buy, but also what you could have done by saving for later. Relevant!
-
China’s AI Workforce: Global Demand for AI Talent Event
By
–
A Virtual CSET Event: China’s AI Workforce – Demand for AI Talent in a Global Marketplace (Thursday, November 10th at 4 pm EST) https://
oodaloop.com/archive/2022/1
1/09/a-virtual-cset-event-chinas-ai-workforce-demand-for-ai-talent-in-a-global-marketplace-thursday-november-10th-at-4-pm-est/
… via @ooda -
ML Insights on Overfitting Apply Beyond Machine Learning Systems
By
–
Excellent post about applying insights from ML (overfitting control) to a much broader class of systems that optimize against an objective: politics, science, orgs, daily life. Underfitting is underrated.
-
AI Dominates TIME’s Best Inventions List for 2022
By
–
0% surprised that @OpenAI Dall-E 2 and @NVIDIA Instant NeRF made it onto the Best Inventions of 2022 list from @TIME
. Artificial intelligence has been its own category on the list since 2020. Expect it to grow. -
MLPerf benchmark requires implementation of mitigations
By
–
MLPerf benchmark needs some of these mitigations https://
x.com/jaschasd/statu
s/1589424193946648576?s=46&t=Yrwu6ciEa83A6HcKopDkNw
… -
Model Capabilities Emerge with Scale: Sentiment Classification Study
By
–
This is a wonderful study on the emergence of model capabilities with scale. It is remarkable to see sentiment classification working with 1 data point, and fascinating to see the effectiveness of chain of thought prompting growing with scale.
-
VIOLA: UT Austin & Sony AI Advance Robot Manipulation Learning
By
–
UT Austin & Sony AI’s VIOLA Object-Centric Imitation Learning Method for Robot Manipulation Outperforms the SOTA by 45.8%
-
ZerO Initialization with Hadamard Transforms for Rectangular Matrices
By
–
And an overview of how ZerO initialization can overcome degeneracy by using Hadamard transforms for rectangular weight matrices
-
Training Degeneracy in Rectangular Weight Matrices Without Hadamard
By
–
This is the training degeneracy that @jiawzhao is talking about if you only do identity initialization and not Hadamard for rectangular weight matrices