Tag
NeuralCert presents a discovery-to-certification framework using neural networks to find and verify extremal mathematical constructions, enabling exact and independently verifiable proofs.
OpenAI is facing criticism for its clumsy handling of mathematical breakthroughs and the formation of an advisory group, causing confusion and frustration among mathematicians.
OpenAI's AI agents have solved the Navier-Stokes existence and smoothness problem, a decades-old mathematical puzzle, challenging traditional artistic approaches to math problem-solving.
This paper proposes Domain-Normalized Multi-Teacher On-Policy Distillation (DN-MOPD) to address unbalanced feedback when merging specialist language models, showing performance improvements on benchmarks like mathematics and instruction-following.
The irrationality of ζ(5) has been proved using an AI-assisted method, and the solution is marked as solved on FrontierMath despite not following the Apéry-style proof as initially sought.
A memorandum between mathematician Hermann Weyl and educator Frank Aydelotte from 1945, likely discussing academic or scientific topics.
Amit Sahai discusses the impact of AI on mathematics, emphasizing the need for more mathematicians to understand AI-generated ideas and collaborate with AI systems to maintain human agency in research.
This paper introduces a framework for LLMs to discover and prove interesting mathematical theorems by optimizing for a metric based on proof difficulty, leading to more novel and useful mathematical knowledge with reduced overlap with existing libraries.
The article delves into the controversial history of Tetris, covering copyright battles over its tetrominoes, their mathematical origins predating the game, and a recent political incident involving a clone removed after legal threats.
The article explains how Fourier analysis can be used to draw complex shapes like llamas using rotating circles, demonstrating the Fourier transform through visual examples.
A study found that disclosed AI use in mathematics-related arXiv papers increased significantly from 1.39% to 14.09% in a six-month period.
The paper introduces FrontierMath Erdős, a benchmark of 68 open Erdős problems for evaluating AI models in mathematics using Lean proof assistant, aiming to address shortcomings in current AI research demonstrations.
OpenAI has formed an independent advisory group of elite mathematicians to guide the company on handling mathematical research and AI interactions, aiming to avoid future reputational crises.
This paper defines intrinsic interestingness for mathematical theorems using proof-to-statement length ratio and trains a 27B model to predict proof difficulty, enabling the generation of more interesting theorems and self-expanding mathematical libraries with reduced overlap with existing knowledge like Mathlib.
OpenAI claimed to solve the Navier-Stokes millennium prize problem, but mathematicians argue the solution dodges the question by only addressing part of the criteria.
Two years ago, AI researchers believed solving a Millennium math problem would take AI 30 years, but recent advancements suggest this timeline may be shorter.
OpenAI has reportedly trained a new AI model that has solved over 100 long-standing open mathematical problems, leading to the formation of an advisory group to assess and coordinate the release of these results.
The article emphasizes that mathematics is an ever-expanding field with countless open problems, and AI is a transformative tool that can advance mathematical research faster than ever before, shifting the focus from individual achievements to collaborative progress.
An open-source autonomous agent has been running for over 50 hours and processed more than 100 million tokens in an experiment to solve the C(25,15,5) covering design problem, with the goal of achieving a breakthrough in mathematics.
A post on X criticizes academia for underestimating AI advancements, pointing to OpenAI's model solving Erdős problems as evidence of major progress.