CS PhD @ Stanford AI Lab, Stanford NLP. Prev Google DeepMind. https://ai.stanford.edu/~kzliuhttps://bsky.app/profile/kzliu.bsky.socialhttps://bsky.app/profile/kzliu.bsky.social/post/3lxd3b6nlcc2kNew paper! We explore a radical paradigm for AI evals: assessing LLMs on *unsolved* questions. Instead of artificially difficult exams where progress ≠ value, we assess LLMs on organic, unsolved problems via reference-free LLM validation & community verification. LLMs solved ~10/500 so far:26 Aug 2025 17:50 +0000at://did:plc:l3juojdkfdbepmj7aqctgcap/app.bsky.feed.post/3lxd3b6nlcc2khttps://bsky.app/profile/kzliu.bsky.social/post/3lbiqad4zik2ngo.bsky.app/AKGJ82V https://go.bsky.app/AKGJ82V22 Nov 2024 00:34 +0000at://did:plc:l3juojdkfdbepmj7aqctgcap/app.bsky.feed.post/3lbiqad4zik2nhttps://bsky.app/profile/kzliu.bsky.social/post/3lbiaxhbmww2yhi21 Nov 2024 20:01 +0000at://did:plc:l3juojdkfdbepmj7aqctgcap/app.bsky.feed.post/3lbiaxhbmww2y