guyga

Joined 6 years ago

Karma

61

Submissions

3

Member since

Jun 19

View on Hacker News

Recent submissions

Comment · 4 years ago

From the paper, the Polish tests are from May 2022, and the training dataset was from January 2022, so this is unlikely.

Minerva: Solving Quantitative Reasoning Problems with Language Models ai.googleblog.com

98 guyga 4 years ago 36 comments

BIG-bench: a collaborative language model benchmark from Google and OpenAI github.com

4 guyga 5 years ago 0 comments