About
About Gradient Damage
Maths you can see. Equations you can hear.
Mathematics you can see. Equations you can hear.
Gradient Damage explores maths through 3D animation, visual explanations and sound. From hidden geometry and number theory to waves, chaos and mathematical physics, the aim is to make abstract ideas feel tangible.
Expect visual maths Shorts, maths-to-sound experiments and stories behind mathematical research. We also explore AI-assisted mathematics, including OpenAI's public research releases, with original sources and a clear distinction between an illustration and a proof.
Start with the Visual Math & 3D Animations playlist. Subscribe for maths that makes you look twice.
English. Occasional strong language.
Not OpenAI
This site and channel are independent. They are not affiliated with, endorsed by or operated by OpenAI. We use OpenAI's name only to say whose published results we are explaining.
How each page is made
- What OpenAI claims is OpenAI's own one-paragraph summary from its public catalogue, quoted as published and attributed to OpenAI.
- Papers link straight to the preprint PDFs in the catalogue, with their dates and abstracts.
- Write-ups are written from the paper and checked against its abstract and statements. Where one is not written yet, the page shows the claim, the papers and the video.
- The animation illustrates the idea of a result. It is not a proof, and the paper is always the authority. Every quantity shown is computed from the real objects; nothing is faked.
- Video pages carry the YouTube description, chapters and, where we uploaded captions, the transcript.
Status of the results
Every result here is a preprint: it has not been through peer review. Some may turn out to contain errors. We report the claims; we don't certify them.
Corrections
If a page misstates a result, say so in the comments of the video and we'll fix the page.
Data and machine readers
Everything on this site is also available as plain data, regenerated with every build: results.json (every catalogue result with its claim, papers, page and video), videos.json (every video with its description, chapters and the results it covers), a summary for AI assistants with a full listing, the sitemap and an Atom feed. Every page carries schema.org structured data. Crawlers of all kinds, including AI assistants, are welcome; quote us with a link.