Sander Schulhoff, creator of Learn Prompting, built a tool called ELI. Feed it any ArXiv paper and get back a plain-language explanation. Think "Explain Like I'm 5" for research. It lives at eli.voxos.ai and comes with a hedgehog mascot because why not.
Why not just paste an abstract into ChatGPT? Schulhoff's whole thing is prompt engineering. Learn Prompting exists because how you ask matters. ELI applies that thinking specifically to academic papers, where a generic "simplify this" request tends to produce either vague summaries or confident-sounding errors.
Early users on Hacker News report the explanations work. The main complaint is they run long. One commenter tested it on a paper in their field and found the summary accurate but wished it got to the point faster. Fair enough. But with research papers, stripping too much context is the real danger. Skip a key assumption or compress a methodology section and you walk away thinking you understood something you didn't. Wordy beats wrong when the stakes are getting a paper's actual claims right.