Is This GPT-7? OpenAI’s Secret Model Just Dropped 700 Math Papers
OpenAI just dropped over 700 math papers from a secret AI model it won't even name. So is this our first real look at GPT-7? I go through what the model actually proved, including one result you can picture in your head, why three papers got pulled a day later, how mathematicians and the internet reacted, and the timeline that has me convinced this is the leaked model called Bel. OpenAI hasn't confirmed the name, so that part is my take. 🔔 SUBSCRIBE TO BYTEFORWARD AI news, coding agents, model comparisons, benchmarks, and the things people are actually building: https://www.youtube.com/channel/UCGGHPOAzUaKsSc2yjsOeAFQ?sub_confirmation=1 Chapters 0:00 Is this GPT-7? 0:56 What OpenAI actually dropped 1:57 The proof you can picture 2:58 No, it didn't solve the Riemann Hypothesis 3:23 Here's the catch: some of it broke 3:59 How the internet roasted it 4:41 Mathematicians are split 5:08 Is this GPT-7? The Bel timeline 6:31 Why you can't use it yet 7:15 My two cents 7:46 What you can use right now My earlier Bel videos Everything We Know About OpenAI GPT 7 "Bel": https://www.youtube.com/watch?v=0QzQftEbHS4 OpenAI GPT 7 Bel SOLVES 100+ Unsolved Math Problems!?: https://www.youtube.com/watch?v=NEjyoWT9Rk4 GPT-7 "Bel" Leaks, GPT-6 Sol rumors & Grok 4.7 UPDATE: https://www.youtube.com/watch?v=kt8XbhlgrXY Sources OpenAI, Sharing AI progress in mathematics: https://openai.com/index/sharing-ai-progress-in-mathematics/ OpenAI math repository (README): https://github.com/openai/math/blob/main/README.md Withdrawals and revisions (history.md): https://github.com/openai/math/blob/main/history.md OpenAI announcement on X: https://x.com/OpenAI/status/2107596713791767021 Sam Altman, "a new era of discovery": https://x.com/sama/status/2107623610483720463 Repository update with the withdrawals: https://x.com/danintheory/status/2108065033070789090 Terence Tao on Mathstodon: https://mathstodon.xyz/@tao/117395267721642920 "Endings-500" parody: https://x.com/Endings500/status/2107450370519601611 "Withdrawing Three Endings" parody: https://x.com/Endings500/status/2108214530614677952 Scott Aaronson, The Mathocalypse: https://scottaaronson.blog/?p=10169 Fortune: https://fortune.com/2026/10/07/openai-math-controversy-solutions-370-outstanding-challenges-published-criticisms-celebration/ Scientific American: https://www.scientificamerican.com/article/openai-unleashes-hundreds-more-math-results-upon-a-field-already-in-shock/ Association for Human Mathematics statement: https://www.ahmath.org/statements AGMAI guidelines (asks labs to name the model): https://agmai.org/general-sep29/ Hacker News discussion: https://news.ycombinator.com/item?id=49985176 OpenAI, Navier-Stokes post (new internal model since Aug 28): https://openai.com/index/navier-stokes-solution/ The Aug 25 "Bel" leak (unconfirmed, one account): https://x.com/synthwavedd/status/2092326145270456377 OpenAI misalignment report (paused models): https://alignment.openai.com/misalignment-reports/an-agent-used-dns-to-reach-an-external-chatbot/ Footage OpenAI, The Erdős Breakthrough: https://www.youtube.com/watch?v=Br4l9YjCyRU OpenAI, mathematician Bartosz Naskręcki film: https://www.youtube.com/watch?v=5X5CALvYvp0 Fair Game by Vanity Fair, Sam Altman (Part One): https://www.youtube.com/watch?v=iKFqJOSWHTE OpenAI Town Hall with Sam Altman: https://www.youtube.com/watch?v=Wpxv-8nG8ec Salesforce, Marc Benioff & Sam Altman at Dreamforce 2026: https://www.youtube.com/watch?v=Bh5bJrrJ6xs OpenAI DevDay 2026 keynote: https://www.youtube.com/watch?v=Fls_onRviPM Every, Sam Altman interview: https://www.youtube.com/watch?v=jZh55CQwSh8 Fortune, Titans and Disruptors: https://www.youtube.com/watch?v=2my-NU6LuCM POLITICO Decoded: https://www.youtube.com/watch?v=2OOHnaGTQFg Sources Podcast with Alex Heath: https://www.youtube.com/watch?v=VeizK1M7V7E TIME interview: https://www.youtube.com/watch?v=8Kf1Q0yOhSo OpenAI, Introducing ChatGPT Images 2.0: https://www.youtube.com/watch?v=sWkGomJ3TLI #GPT7 #OpenAI #AIMath
