John Yang (@jyangballin) 's Twitter Profile
John Yang

@jyangballin

CS PhD @Stanford 🌲 | SWE-bench/agent/(collab w/ me!) 🤖 | Prev. @princeton_nlp 🐯, @Berkeley_EECS 🐻

ID: 616998786

linkhttps://john-b-yang.github.io/ calendar_today24-06-2012 08:31:40

230 Tweet

2,2K Followers

573 Following

John Yang (@jyangballin) 's Twitter Profile Photo
John Yang
@jyangballin

3 months ago

SWE-bench + OpenAI = 𝗦𝗪𝗘-𝗯𝗲𝗻𝗰𝗵 𝗩𝗲𝗿𝗶𝗳𝗶𝗲𝗱! A subset of 500 problems w/ a theoretical ceiling of 100% performance curated from human annotations Really excited to finally address the mystery of human performance on SWE-bench!