According to the Stanford AI Lab blog, CausalSmith is an agentic framework that automates theoretical causal inference research by using Lean to verify proofs and an audit step to align formal statements with paper claims.

According to AnthropicAI, Claude produced the first Lean formalization of Fermat’s Last Theorem with 13M lines, validating 29k lemmas across mathlib.

According to emollick, OpenAI’s Astra released 10 formal math proofs with Lean and CoT, signaling hard-to-feel capability gains across fields.

According to @emollick, OpenAI’s Astra produced 10 formal math proofs with Lean and CoT, showing rapid gains in STEM reasoning and low API costs.

According to @gdb, OpenAI’s Astra delivered 10 new proofs with Lean certificates across algebra, packing, and complexity at ~$2000 each.