🎉 I'm giving away a free spot in my next Applied Conformal Prediction course! 🏆
To enter, tweet why you want to join and use #mavenconformalprediction.
Don’t forget to tag me! Best post submitted before October 31st wins! 🚀
Looking forward to your entries!
Typical academic paper:
Theory A proposes everything is A
Theory B proposes everything is B
Our groundbreaking theory proposes things are sometimes A and sometimes B
If time is the coin of life, how do people spend it?⌛️
🚨 Our new @PNASNews article “The Global Human Day” integrates time-use data from hundreds of datasets to provide a global glimpse of what people are doing on earth. Work led by William Fajzel 🌐 1/
https://t.co/7qANDhGqsp
@RobertHaisfield If I remember 12 correctly, it's that heavy lifts require ramp-up time, which can be reduced with effective use of intermediate packets
Elegant explanation of why starting prompts with "you are a causal inference assistant" improves results: the distribution of words in documents about causal inference differs from that of the complete training set, so conditioning on that focuses the LLM response.
You're invited: Join a @Metaculus-hosted technical talk exploring the causal reasoning abilities of LLMs, featuring @Microsoft Principal Researcher @amt_shrma and Sr. Principal Researcher Emre Kichirin. 370+ people have RSVP'd so far. Bring your questions! https://t.co/342htRc5n9