Noam Brown Laments the Massive Changes in AI Math Benchmarks Over Two and a Half Years
AI summary
Noam Brown recalled that just 2.5 years ago, AI labs would still prominently showcase the performance of their frontier LLMs on the elementary school math benchmark GSM8K, and said he is very much looking forward to the progress in another 2.5 years.
View original postIt just never gets old https://t.co/H9rsmgj99U
View original
It just never gets old https://t.co/H9rsmgj99U
Body
Only 2.5 years ago, AI labs would highlight their frontier LLM performance on grade school math benchmarks (GSM8K). I can't wait to see where we'll be in another 2.5 years.