@superlopuh if by "this sort of testing" you mean generative testing more broadly, yes it's seen spectacular successes with compilers, and indeed compiler testing is where most of the cutting edge research on it goes on. Compilers are reliably full of bugs that are hard to catch any other way
Also for python in particular there's Hypothesis, which I wrote so am entirely unbiased when I say yes every python project should be tested with it
@superlopuh it depends a bit on what the question means.
Antithesis for compiler testing? Probably not. Antithesis is spectacular if you have Antithesis shaped problems (currently mostly distributed or highly concurrent systems) and Antithesis shaped budgets (currently very high. We want to offer more affordable options but can't yet) and not so great outside that range. I'd not expect a compiler to have either
But...
Anyway I partly mention this because work asked me to let you know that the second BugBash, our conference on software reliability, is coming up. https://bugbash2026.antithesis.com/
I figure enough of you know a thing or two about computers that this might be interesting. I didn't make it to (/know about at the time) the first one last year, but I've watched some of the talks and they're great. https://www.youtube.com/playlist?list=PLh4UhOpNuTJM5A_Ztsaz6OHQccdYowufH
I keep forgetting to mention this in public, but I work at https://antithesis.com/ now. Joined back in November. It's been going great so far.
It's very funny, and really remarkable how often it turns out to be the case that Antithesis and Hypothesis are opposite (but complementary) to each other, e.g. on a power vs convenience curve. I'm hopefully there to help bridge some of the gap between the two.
