Tag
A claim surfaced at NeurIPS that the Riemann Hypothesis has been solved, prompting a request for an extremely simple, bottoms-up plain-English explanation of the math and why the result would be significant.
Levent Alpöge claims to have proven the Riemann Hypothesis (and the Generalized Riemann Hypothesis) with assistance from Claude, including a Lean formalization of the proof, though independent verification is still pending.
An Anthropic employee has released a PDF claiming to present a general proof of the Riemann Hypothesis, one of the most famous unsolved problems in mathematics. The claim has not yet been verified by the mathematical community.
David Harvey and Joris van der Hoeven's paper presenting a groundbreaking O(n log n) integer multiplication algorithm, breaking the long-standing complexity barrier for multiplying large integers.
Leading number theorists are growing increasingly confident in OpenAI's AI-generated proof of the quasi-Riemann hypothesis, with Terence Tao publishing an exposition of a related simplification by Ben Green, marking a notable shift toward mathematicians independently evaluating AI-produced frontier mathematics.
The author factored the RSA-896 challenge number using the Claude AI on September 19, 2026, providing the prime factors and showcasing AI's role in cryptography.
Researchers from Cognition used the AI agent Devin to factor RSA-260, setting a new record in the RSA Factoring Challenge and demonstrating the potential of autonomous software engineering in cryptographic research.
An elliptic curve with rank at least 30 has been discovered, setting records in conductor, naive height, and other invariants on the Elliptic Curve Rank Leaderboard.
Anthropic's unreleased research version of Claude improved the lower bound for the fraction of Riemann zeta function zeros satisfying the Riemann hypothesis from 41.6% to 67.2%, demonstrating rapid progress in AI mathematical capabilities.
Andrew Wiles reflects on his journey to proving Fermat's Last Theorem, focusing on the pivotal moment of final insight and the fulfillment of a childhood dream.
The author claims to have solved six open problems posed by the mathematician Paul Erdős within a span of five days.
A new formal theorem, verified in Lean, shows that for thresholds tending to infinity, almost every positive integer falls below the threshold within 436 log N Collatz steps, strengthening Terence Tao's earlier result with explicit bounds and natural density.
Number theorists discovered a counterexample to the Jacobian conjecture in three-dimensional space using the AI model Fabel. After 87 years, the conjecture was disproved, but the planar version remains open.
Researchers used 20 parallel Codex accounts to solve 20 Erdős problems, including a formal proof of Erdős problem #123 in number theory using Lean.
This paper trains a two-layer transformer encoder to classify rational elliptic curves by rank from Frobenius traces, achieving >99% accuracy. Mechanistic interpretability reveals the model learns the Mestre-Nagao heuristic and concentrates attention on prime positions, demonstrating that transformers can learn number-theoretic algorithms.