Note that this refers to ML as in machine learning and not the ML language family, as I hoped.
My mind still unabbreviates "ML" into "Machine Language" before anything else, only because it was the first of these three for me.
Research papers on ML in Compilers
11–20 of 27 posts
Re: Research papers on ML in Compilers
#12There is a mathematical description of programming compilers I read once (think it was a survey paper) that I found very clear but I've never been able to find since. Basically defined a compiler as a type of function. Any good suggestions?
Perhaps this one from Graham Hutton?
“A parser for things
Is a function from strings
To lists of pairs
Of things and strings”
Re: Research papers on ML in Compilers
#13My research (ML for binary function recognition) uses obfuscation and diverse compilation for data augmentation.
Intuitively, obfuscation is a form of anti-optimization. My hope is that it's differentiable, so perhaps given enough knowledge of obfuscation, an ML model can make corresponding optimizations to "undo" them, then for more performance, apply those same un-obfuscations to code that hasn't already been obfuscated.
Re: Research papers on ML in Compilers
#14I'd be more interested in the reverse, that is compiler and PL techniques applied to machine learning.
You might be interested in this: https://github.com/merrymercy/awesome-tensor-compilers
Re: Research papers on ML in Compilers
#15It's interesting to add the work about Deepmind AlphaDev to this list.
Re: Research papers on ML in Compilers
#16My research (ML for binary function recognition) uses obfuscation and diverse compilation for data augmentation. Intuitively, obfuscation is a form of anti-optimization. My hope is that it's differentiable, so perhaps given enough knowledge of obfuscation, an ML model can make corresponding optimizations to "undo" them, then for more performance, apply those same un-obfuscations to code that hasn't already been obfus…
I think this is a great line of research as it may solve the reasoning portion of AGI. The difficulty is that you would want to prove equivalence. LLMs can output stuff but for programs to execute you need to output correct stuff.
Re: Research papers on ML in Compilers
#17Related C2 wik page:
Re: Research papers on ML in Compilers
#18Note that this refers to ML as in machine learning and not the ML language family, as I hoped.
There are so many flavors of machine learning that we should probably just refer to whole field as XML.
Re: Research papers on ML in Compilers
#19There is a mathematical description of programming compilers I read once (think it was a survey paper) that I found very clear but I've never been able to find since. Basically defined a compiler as a type of function. Any good suggestions?
Perhaps this one from Graham Hutton? “A parser for things Is a function from strings To lists of pairs Of things and strings”
hmmmm maybe something like page 184 here https://apps.dtic.mil/sti/pdfs/ADA327435.pdf
Re: Research papers on ML in Compilers
#20There is a mathematical description of programming compilers I read once (think it was a survey paper) that I found very clear but I've never been able to find since. Basically defined a compiler as a type of function. Any good suggestions?
Perhaps this one from Graham Hutton? “A parser for things Is a function from strings To lists of pairs Of things and strings”
Thanks for sharing this. I became a bit of a fan of Dr Hutton after I followed his Functional Programming in Haskell classes on Youtube.