Two papers used GPT6 Astra to find counterexamples to a plasma physics question open since 1967, showing stable smooth 3D plasma configurations can exist without symmetry.
Counterexamples to Grad's conjecture in plasma physics, with a Lean 4 formalization of the main result.
Updated EU status checked Last shown
Plasma conjecture disproof at a glance
- What is Plasma conjecture disproof?
- Plasma conjecture disproof is a research. Counterexamples to Grad's conjecture in plasma physics, with a Lean 4 formalization of the main result.
- Does Plasma conjecture disproof work in the EU?
- Available in the EU. The material is a publicly accessible GitHub repository and arXiv paper with no geographic restrictions; no EU-specific evidence was found beyond that. Last verified 14 Nov 2026.
- How much does Plasma conjecture disproof cost?
- Free. Public GitHub repository; no pricing. License not specified in the provided material.
- Where was Plasma conjecture disproof shown?
- In 1 creator video by AI Search. Most recently by AI Search on 4 Oct 2026, at 25:06 in “Gemini 4, GPT 6.1, Plasma conjecture, Napoleon code solved, OpenAI Dots, Claude Sonnet 5.5: AI NEWS”.
Sources: the vendor's own pages (linked above) and creator videos at the linked timestamps. EU status is information, not legal advice; see how we check.
"Counterexamples to Grad's conjecture" is a mathematical result by Javier Gómez-Serrano, Lukas Liehr and Mitchell A. Taylor, concerning magnetohydrostatic (MHS) equilibria in plasma physics. The GitHub repository lukasliehr/Grad-Conjecture contains a Lean 4 formalization of the paper's main theorem. The theorem constructs smooth families of embedded solid tori carrying smooth MHS equilibria with only a finite cyclic rotational symmetry, i.e. without continuous symmetry. The fields satisfy B×(∇×B)+∇P=0 and ∇·B=0, with B tangent to the boundary.
The formalized statement says that for any cell length L>0 there is a compact interval of parameters and an integer N0 such that for every N≥N0 the family has a pressure with a single round critical circle of radius NL, a magnetic field vanishing exactly on that circle, and a stabilizer equal to the cyclic group of N rotations. The family gives a continuous injective curve in the moduli space of embedded configurations whose interior members are non-isolated.
The Lean library covers Nash–Moser iteration, smooth parameter dependence, Fourier analysis, analytic weights, Banach calculus, topology and the MHS application. Two showcase files state the result: one imports only Mathlib and uses two sorry placeholders, and the other supplies full proofs. The proved result depends only on the standard axioms propext, Classical.choice and Quot.sound. A YouTube creator reported that the work used a GPT-6 model to find the counterexamples, but this is not confirmed in the provided repository text. This is a research artifact, not a product.
- Lean 4 formalization of the main theorem of 'Counterexamples to Grad's conjecture'
- Smooth MHS equilibria on solid tori with only cyclic N-fold symmetry
- Proof library covering Nash–Moser iteration, Fourier analysis and Banach calculus
- Showcase files with and without proofs, checked by a Comparator suite
- Depends only on standard axioms propext, Classical.choice, Quot.sound
As seen on YouTube
Watch the moment: the exact second top creators show it. Videos load only when you press play, via youtube-nocookie.com.
Similar AI tools
Closest neighbours by meaning, not just by category.
Hydrogen 1.0
Model · DeepPrinciple
753B-parameter open-weights scientific-discovery model from DeepPrinciple, post-trained from GLM 5.3.
OpenAI scientific discoveries
Research · OpenAI
OpenAI's announced AI-produced results: a local smoothing proof, an ETH refutation and a QCD result, with Lean verification.
OpenAI math breakthroughs
Research · OpenAI
OpenAI repository of 700+ math manuscripts and Lean proofs generated by an unreleased internal model.
OpenAI Navier-Stokes solution
Research · OpenAI
OpenAI's announced AI-generated finite-time blowup result for Navier–Stokes, with a writeup and Lean formalization.
AsymFlow
Model
Rank-asymmetric flow method for pixel-space image generation, incl. AsymFLUX.2 klein finetuned from FLUX.2 klein 9B.
Room temperature magnetic semiconductors discovery
Research · Vals AI
Vals AI report on AI agents using DFT simulations to find two room-temperature antiferromagnetic semiconductor candidates.
