3 ms·Large language models beat undergrads on tests of reasoning via analogy3 points by LinuxBender 3y ago