Pop Culture Pulsar: Origin Story of Joy Division’s Unknown Pleasures Album Cover
Great dig into the CP1919 pulsar signal plot that was used for “Unknown Pleasures”:
This plotting of sequences like this, it started just a little bit earlier when we were looking at potentially drifting subpulses within the major pulse itself. So, the thought was, well, is there something like this peak here, which on the next pulse moves over here, and then moves over here, and over there. Actually, would be moving this way in that case – either way. I think Frank Drake and I published a paper in Science Magazine on exactly that issue – suggesting there might be drifting subpulses within the major pulse, which would then get back to the physics of what was causing the emission in the first place. So, then the thought was, well let’s plot out a whole array of pulses, and see if we can see particular patterns in there. So that’s why, this one was the first I did – CP1919 – and you can pick out patterns in there if you really work at it. But I think the answer is, there weren’t any that were real obvious anyway. I don’t really recall, but my bet is that the first one of these that I did, I didn’t bother to block out the stuff, and I found that it was just too confusing. So then, I wrote the program so that I would block out when a hill here was high enough, then the stuff behind it would stay hidden. And it was pretty easy to do from a computer perspective.
(tags: design joy-division music science physics pulsars astronomy cp1919 dataviz)
moyix/gpt-wpre: Whole-Program Reverse Engineering with GPT-3
This is a little toy prototype of a tool that attempts to summarize a whole binary using GPT-3 (specifically the text-davinci-003 model), based on decompiled code provided by Ghidra. However, today’s language models can only fit a small amount of text into their context window at once (4096 tokens for text-davinci-003, a couple hundred lines of code at most) — most programs (and even some functions) are too big to fit all at once. GPT-WPRE attempts to work around this by recursively creating natural language summaries of a function’s dependencies and then providing those as context for the function itself. It’s pretty neat when it works! I have tested it on exactly one program, so YMMV.
(tags: gpt-3 reverse-engineering ghidra decompilation reversing llm)