Meta's Muse Spark solves six open math problems

Meta published six papers co-written by mathematicians and the Muse Spark model, claiming it solved six open problems via the ordinary meta.ai chat.
Meta's Muse Spark has solved six open mathematical problems.
Meta published six papers written by mathematicians together with the model. According to the company, researchers used Muse Spark 1.1 and 1.2 in Thinking Mode through the ordinary meta.ai chat, without a special harness.
Notably, each paper even indicates which fragments were written by humans and which by AI.
The results span several areas of mathematics: probability theory, differential equations, optimization, group theory, algebra, and arithmetic physics.
For example, the model disproved Kida's 2024 conjecture: a semi-abelian group does not have to be monomial. It was also proved that radially symmetric solutions with negative energy for the biharmonic nonlinear Schrödinger equation collapse in finite time, a question that had remained open since 2015.
However, in September another group of researchers also disproved Kida's conjecture, also using an AI agent. Meta says its results were obtained independently.
Why it matters
AnalysisThis suggests frontier models can contribute to genuine mathematical research, not just code or text generation. The explicit human/AI attribution may become a useful norm for academic publishing.
Discuss in community
Share your questions and insights with developers
Related in Ecosystem
Practical guides, core terminology, and agent skills
Llama Family (Meta Llama)
A series of foundational open language models from Meta (Llama 3, 3.1, 3.3) that have become the industrial standard for the Open Weights ecosystem, local AI, and enterprise fine-tuning.
Anthropic mathematician claims possible proof of Riemann hypothesis with Claude
Levent Alpoge says collaborative work with Claude may have led to a proof of the Riemann hypothesis.
Hermes Profile: OSINT Researcher & Deep Web Intelligence
Hermes open-source intelligence (OSINT) profile: multi-hop web investigation, primary source verification, competitive intelligence, and citation-backed synthesis.
NotebookLM + Perplexity for work, study and income
What is NotebookLM and Perplexity