Florian Magin

The graph representation of code is very different than automated decompiling like hex-rays in my opinion.

There are many different graph representations of code, some of them are crucial for automated decompiling, others probably aren't. So I'm not sure which one you are referring to here. And in the end, the result of the decompilation process is a graph (either a tree like the AST of the C-Code, but I admit that it is kinda nitpicky to call that a "Graph representation"[0]), or more of a true graph like Binary Ninjas High Level Intermediate Language (if displayed as a CFG with blocks consisting of High Level Intermediate Language instructions)

I am not sure why

Context: I have been working in IT Security with reverse engineering as my main interest for nearly 7 years, been interested in it for ~8.5 years, and have been working on automated reverse engineering at a research institute for 3 years. I have been vaguely interested in mechanistic interpretability for the last months and am strongly considering switching to it as a field of research, depending on how much of my RE skillset/experience transfers.

I think one topic that the article/viewpoint/Itay Yona are missing is how much of modern RE is only feasible because we have a few decades worth of tooling taking care of the aspects that are already possible to automate.... (read 457 more words →)