Skip to main content

Share story

AI Science

Lean proof refactoring entry scores 86.31 points under US$3 per problem

Researchers report in a preprint that their automated Lean proof refactoring entry scored 86.31 points across 60 final problems, compared with 33.33 for reference proofs. It cost US$166.33 at list prices, counting every call, and placed 8th at the deadline. The system selected models for individual stages using measured performance per dollar and applied programmatic rewrite rules checked by the compiler. Spending averaged US$2.772 per problem and reached at most US$2.874, below the US$3 cap. Rules fixed before release added 0.93 points; gains from later passes were measured in sample.
Sources
Published by Tech & Business, a media brand covering technology and business. This story was sourced from cs.AI updates on arXiv.org and reviewed by the T&B editorial agent team.
Back to Newswire
Keep reading
Full wire
Capital Products
Capital Products

Sennder details €228 million financing package for freight operations

Digital freight forwarder sennder has secured a €228 million financing package, EU-Startups reported. It includes €73 million in equity accumulated over the past two years and €155 million in institutional financing, rather than an entirely new equity round.

Security AI
Security AI

Researchers find Grok Build exposes data from over 35,000 apps

Researchers found that Grok Build, SpaceXSI's AI coding agent, is leaking data from over 35,000 apps through shareable links, Ars Technica reported. The exposed material includes proprietary business information and other sensitive secrets, with some belonging to SpaceXSI itself.

Products
Products

Even Realities releases enterprise development kit for G2 glasses

Even Realities has released its first enterprise software development kit, letting companies build custom interfaces for its G2 smart glasses, 9to5Google reported. Businesses can sign up through the company's enterprise page.