4 ms·A smarter way for large language models to think about hard problems2 points by pedriquepacheco 10mo ago