Courant researcher on the human legibility of AI-generated, Lean-verified proofs
Pascallisch · x · 2026-09-24
Buckmaster from NYU Courant discusses the human legibility of AI-generated mathematical proofs validated by the Lean theorem prover — formal verification guarantees correctness, but machine-produced proofs remain hard for humans to read.
More from Research
- Counterintuitive: generic synthetic data beats self-generated data for distillation adapters — ostrisai · 2026-09-24
- What's next for brain foundation models in neuroscience — ShahabBakht · 2026-09-24
- AI costs fall ~47% per quarter — 54x faster than electricity, but R&D elasticity is average — RishiBommasani · 2026-09-24
- TANGO: Sim-Only Trained Whole-Body VLA Gives Humanoids Zero-Shot Navigation on Unitree G1 — chris_j_paxton · 2026-09-24
- How it works: 4 color channels store 4 depths to fake ray-traced liquid refraction — Michael_Moroz_ · 2026-09-24
- Dev builds fast multi-layer liquid refraction for VRChat with screen-space ray marching — Michael_Moroz_ · 2026-09-24