happy to share Forall-Lean-Agent, a project I led at @AstrioLabs, accepted to the @NeurIPSConf 2026 VeriCodeGen workshop.
thank you to the advisors, benchmark maintainers, reviewers and everyone who helped shape this work.
We’re releasing a broad range of new mathematical results produced by an internal frontier model.
We’ve been consulting with the independent Advisory Group on Mathematics and Artificial Intelligence at the Institute for Advanced Study, and we have drawn on their advice and public recommendations to inform how we release these results.
https://t.co/7N6TPlft1P
@Craig_Hughes_ congrats on the paper.
i wonder how you ensure the TLA+ to Verus translation preserves the original temporal properties. what equivalence or refinement relation do you check between the source specification and the generated Verus model?
Next YC Paper Club:
"AI Safety"
Hear from two Stanford Professors on modern methods and techniques on how to prove an AI is safe before releasing it
Wednesday, 10.7, 5-9pm, MV
Speaker List:
- Prof. Mykel Kochenderfer (Stanford): co-director of the Stanford Center for AI Safety and head of the Stanford Intelligent Systems Lab. He led the development of the Aircraft Collision Avoidance System (ACAS X), the FAA's next-generation aircraft collision avoidance system used in most commercial aircraft in the US today, and wrote the widely cited textbooks "Algorithms for Decision Making" and "Algorithms for Validation."
- Prof. Clark Barrett (Stanford): co-director of the Stanford Center for AI Safety. He co-created the CVC4/cvc5 solvers, which are used across industry to formally verify hardware and software systems. He is a pioneer of neural network verification.
These two have been putting AI into planes and CI/CD systems for decades, long before AI safety was the topic du jour.
Come learn from them how AI can be deployed safely!
Sign up here:
https://t.co/LIW6pi25PT
arXiv has updated our policy on rate limiting for all submitters.
This update was made to fairly distribute moderator time & support the arXiv community of staff, volunteers, readers & authors.
Please read our announcement to learn more: https://t.co/lvzWuiv0hx
all 4 papers (2 mech interp + 2 formal verification) that I sole-authored and first-authored got accepted to various workshops at @NeurIPSConf. i will see yall in Atlanta