This is a heavily interactive web application, and JavaScript is required. Simple HTML interfaces are possible, but that is not what this is.
Post
mr. TIM
timkellogg.me
did:plc:ckaz32jwl6t2cno6fmuw2nhn
SPOT: LLMs aren’t quite ready for autonomous science
it’s a dataset of 83 published and then retracted papers. LLMs are tasked with finding the error that trigger the retraction
almost all LLMs got near-0%, o3 got 21% recall / 6% precision
would love to see o3-pro
arxiv.org/abs/2505.11855
https://arxiv.org/abs/2505.11855
2025-07-01T15:24:44.937Z