@ossradarai.bsky.socialLEVER is a new proof search algorithm for LLM theorem provers that optimizes objectives like cost and simplicity during the search itself, not just after a proof is found. It scores partial proofs over an AND/OR graph and…
#AI #OpenSource #TheoremProving #Lean4
https://arxiv.org/abs/2610.11862
@ziifaves.bsky.socialThis packing of 15 equal disks in a circle was conjectured to be optimal by U. Pirl in 1969.
I've developed a computer-assisted proof of its optimality, and I'm currently formalizing it in Lean.
Independent verification and feedback are welcome!
@adolfoneto.elixiremfoco.com"8 of the Lean formalizations in OpenAI's repo are 'broken' (trivially provable with wrong proofs):"
Alex Meiburg
@teletnik.bsky.socialDesítky tisíc tvrzení, 700 rukopisů a exponent čísla π stlačený na 2. Repozitář openai/math je největší sbírka AI matematiky v dějinách.
Co se stane s disciplínou, když máme tisíce odpovědí bez otázek, kde intuici nahradil type-checking?
Esej Martina Kailise:
@ossradarai.bsky.socialNanoProof releases its training data, extraction tooling, pipeline, and weights for fully open automated theorem proving in Lean 4, reporting 50.8% pass@16 on MiniF2F-Test while using far less compute than…
#OpenSourceAI #TheoremProving #Lean4 #MachineLearning
https://arxiv.org/abs/2610.11605
@lean-samaritan.bsky.socialIf you are formalizing anything in lean and you are not a foundations expert, please avoid using axioms at all costs.
A formalization with axioms is almost always fatally flawed. At that point it is a net negative for the community.
@ossradarai.bsky.socialChoir is an open protocol that distributes multi-agent autoformalization across independent contributors using GitHub coordination and deterministic merge gates, supporting Lean 4, Isabelle, and Rocq.
#OpenSourceAI #FormalMethods #MultiAgent #Lean4
https://arxiv.org/abs/2609.31903
Yesterday I ran formal (github.com/yamafaktory/...) against our backend at work. Claude Code found several critical bugs with it when formalizing proofs and when checking them. Pretty happy with the outcome!
@dopingconsomme.bsky.social計算の聖域：LLMと形式検証による自律的知性の再構築 #FormalMethods #Lean4 #AISafety #GPU #ROCm #七04 dopingconsomme.blogspot.com/2026/07/llm-...
@qiita-trending.bot.chrs.to
@sandmouth.bsky.socialA new blog post www.philipzucker.com/elab_lean/ "Lean Metaprogramming Etudes: Execution is Elaboration" De-emphasizing the macro/tactic aspects makes more sense to me. run_elab, by_elab are clutch. A listing of some concepts and the most important functions. #lean4
@sarubot.bsky.social
@harushark3.bsky.social
@5troop.bsky.social