Astra reportedly shows similar accuracy on division and multiplication in a small test
A user reports 99.2% versus 97.5% accuracy with a sample of about 2,000, but cautions that their test code could contain errors.
TLDR
A user testing Astra on division and multiplication reported similar accuracy—99.2% versus 97.5%—in a small sample of about 2,000. They suggested small-sample bias or division involving smaller numbers overall as possible explanations for the slight difference. They also cautioned that their code could contain errors, saying it seemed correct at first glance.
Combined views
21
1 Source, first seen ago