(adamj.eu)

378 points rbanffy | 1 comments | 09 Dec 25 20:33 UTC | HN request time: 0.368s | source

Show context

mentos ◴[10 Dec 25 08:47 UTC] No.46215516[source]▶

Given that Python tends to produce fewer hallucinations when generated by LLMs I wonder if former Django developers using AI tools are secretly having a blast right now.

replies(4): >>46215588 #>>46215786 #>>46216131 #>>46216252 #

m_ke ◴[10 Dec 25 10:42 UTC] No.46216252[source]▶

>>46215516 #

What a lot of people don’t know is that SWE-bench is over 50% Django code, so all of the top labs hyper optimize to perform well on it.

replies(1): >>46223642 #

1. kristianp ◴[10 Dec 25 20:51 UTC] No.46223642[source]▶

>>46216252 #

I know python is more prevalent in SWE-Bench than any other language, but more than 50% django sounds like a big stretch. Citation?

Edit, it's about 37%, and python-only. https://arxiv.org/pdf/2310.06770v3

↑

Django: what’s new in 6.0