AI Lean proofs can quietly rely on sorry or bypass the kernel via native_decide and unsafe
gklambauer · x · 2026-10-13
Mathematician Gregor Kambauer discusses pitfalls of AI-generated Lean proofs. Beyond known issues, two more failure modes were raised:
- Proofs may rest on sorry or hidden axioms, so they compile without actually being complete.
- The checker itself can be the weak link: features like nativedecide, unsafe, and metaprograms can circumvent Lean's kernel, meaning "it compiles" doesn't guarantee the proof is sound.
The takeaway: when using AI for formal proofs, never trust compilation status alone—watch for paths that bypass kernel verification.
Related event: Researchers Detail How AI-Generated Lean Proofs Can Go Wrong(3 posts)→
More from Research
- Chollet: Ideal AI-for-Science Outcome Is Millions of AI-Aided Scientists, Not a Few Rich Labs — fchollet · 2026-10-14
- Feeling a dipole's electric field along an ellipse: 2.08kq at the end, 0.13kq on top — TinfoilTricorn · 2026-10-14
- Alibaba's Representation Guidance: plain pixel-diffusion Transformers beat bigger models — bdsqlsz · 2026-10-14
- Visualizing the Fourier series of a sawtooth wave, with Gibbs overshoot hitting ~9% — TinfoilTricorn · 2026-10-14
- Gemma Embedding 2 hits SOTA on Vietnamese speech after 3,000 hours of audio, 2 hours on 8xH100 — stablequan · 2026-10-14
- KILVO: robust multimodal odometry for humanoid robots fusing encoders, IMU, LiDAR and vision — rsasaki0109 · 2026-10-14