Verified Reasoning: NeurIPS Happy Hour w/ Axiom, B Capital, and Madrona
Verified Reasoning: NeurIPS Happy Hour with Axiom × B Capital × Madrona
If you are into reinforcement learning / reasoning, formal verification, and AI for mathematics, this is the event for you!
Take a breather from the NeurIPS bustle and join us for an invite-only gathering of AI researchers, engineers, and founders. We’re bringing together a sharp group of builders, from frontier model labs to applied ML teams, to unwind, connect, and trade ideas over drinks and light bites.
Spend the evening in unscripted conversations with teams from Axiom, B Capital, and Madrona, plus technical guests from across the NeurIPS community. No stage, no pitch, just open discussions about the experiments and breakthroughs shaping next-generation models, agentic systems, and production ML.
We’ll also create space for deeper conversations around Axiom’s work at the frontier of mathematical discovery. AxiomProver recently proved Erdős problems #124 and #481 in Lean, a 100% verifiable language, demonstrating meaningful progress in automated reasoning. From automated theorem proving and autoformalization to self-play conjecturing, Axiom’s approach grounds reasoning in programming languages and uses AI to enable new knowledge generation and verification at scale.
Join us to:
Connect with researchers working on automated reasoning, agent systems, large-scale training, and applied ML
Trade ideas on advances in formal methods, math-native AI models, and high-integrity reasoning systems
Swap insights on the most compelling NeurIPS papers—from proofs to agents to scaling laws
Enjoy cocktails, light bites, and great conversation
To keep the event high-signal, registrations will be reviewed and confirmed prior to the event.