- Today
- Holidays
- Birthdays
- Reminders
- Cities
- Atlanta
- Austin
- Baltimore
- Berwyn
- Beverly Hills
- Birmingham
- Boston
- Brooklyn
- Buffalo
- Charlotte
- Chicago
- Cincinnati
- Cleveland
- Columbus
- Dallas
- Denver
- Detroit
- Fort Worth
- Houston
- Indianapolis
- Knoxville
- Las Vegas
- Los Angeles
- Louisville
- Madison
- Memphis
- Miami
- Milwaukee
- Minneapolis
- Nashville
- New Orleans
- New York
- Omaha
- Orlando
- Philadelphia
- Phoenix
- Pittsburgh
- Portland
- Raleigh
- Richmond
- Rutherford
- Sacramento
- Salt Lake City
- San Antonio
- San Diego
- San Francisco
- San Jose
- Seattle
- Tampa
- Tucson
- Washington
Eclipse Corrects Errors in Autoformalization Research Release
Founder Neel Somani reflects on insights from AI-driven exploration of unsolved Erdős problems.
Mar. 17, 2026 at 1:40am by Ben Kaplan
Got story updates? Submit your updates here. ›
Eclipse, a blockchain platform founded by computer scientist Neel Somani, has issued a correction to a previous news release about the company's research into autoformalization and the use of advanced AI models to tackle unsolved mathematical challenges. The corrected release provides more details on the "GPT-Erdos" project, which employed tools like GPT-5.2 Pro and Deep Research to explore the potential of AI in mathematical discovery. Somani discusses the insights gained, including the challenges of defining novelty and "interestingness" when AI-generated solutions diverge from established approaches.
Why it matters
Somani's work highlights the profound implications of autoformalization, the process of converting human-readable proofs into machine-verifiable formats. As AI systems increasingly engage in this process, they are prompting the scientific community to reconsider long-held notions of novelty, progress, and rigor in mathematical research. The insights from the GPT-Erdos project could have far-reaching impacts, from redefining software reliability standards to transforming the nature of academic inquiry.
The details
The GPT-Erdos project organized a team of undergraduate students to apply advanced AI models to unresolved Erdős problems, with the goal of not only solving mathematical challenges but also uncovering the deeper principles that guide human research. The results were striking, yielding accepted solutions, partial progress, and rediscoveries of previously undocumented findings. However, Somani notes that the true value of the initiative lies in the insights it provided into the informal, often hidden principles that govern mathematical discovery.
- The GPT-Erdos project was launched in early 2026.
- The corrected news release was issued on March 5, 2026.
The players
Neel Somani
A Berkeley-educated computer scientist and the founder of the blockchain platform Eclipse, who spearheaded the GPT-Erdos project.
Terence Tao
A leading mathematician whose views on the novelty of AI-generated results are cited in the release.
Eclipse
A blockchain platform founded by Neel Somani, which organized the GPT-Erdos project to explore the potential of AI in mathematical discovery.
What they’re saying
“The fusion of artificial intelligence and pure mathematics is accelerating, evolving from computational assistance to enabling creative breakthroughs.”
— Neel Somani, Founder, Eclipse (PR Newswire)
“When AI generates solutions without a clear historical lineage, it challenges the human desire for clean definitions of "newness," revealing the often messy reality of mathematical derivation.”
— Neel Somani, Founder, Eclipse (PR Newswire)
What’s next
Somani and the Eclipse team plan to continue exploring the implications of autoformalization, with a focus on developing tools that can measure the "closeness" of AI-generated proofs to completion. This could transform AI from a binary checker into a true collaborator in the process of mathematical discovery.
The takeaway
The GPT-Erdos project has revealed the profound impact of autoformalization on the practice of mathematics, challenging long-held notions of novelty, progress, and rigor. As AI systems become increasingly adept at generating solutions, the scientific community must grapple with how to define and assess the value of these contributions, paving the way for a new paradigm of inquiry.
San Francisco top stories
San Francisco events
Mar. 17, 2026
Joe Klocek & FriendsMar. 17, 2026
Clinton Kane - 4350 Live with Julian Ray




