News for hackers. Stuff that matters.
Re:Ask HN: Those with success using GPT-4 for programming – ... by acqq on Monday May 22, 2023 @07:21PM (#36037957)
But 3000003 is exact in fp32, as well as it is 3000002.5, 3000002.75 before it or 3000003.25, 3000003.5, 3000003.75 after it?

https://www.h-schmidt.net/FloatConverter/IEEE754.html

And AFAIK the LLMs, lacking the ability to actually calculate, can't check which numbers are such?

Re:Ask HN: Those with success using GPT-4 for programming – ... by siboehm on Monday May 22, 2023 @07:33PM (#36038042)
Ha, you're right. I should've used 30,000,000. The answer it gave regarding how to do "ties to even" was still correct though: "Evenness" is decided by the least significant bit of the mantissa, not be the "evenness" of the decimal representation.