AILLM reportedly taught to find interesting theorems without human guidance
A post introducing the work claims 4.3× higher “interestingness,” with a quantitative notion of that quality and a self-expanding discovery loop.
TLDR
The author argues that finding interesting theorems without human guidance is a new bottleneck as LLMs prove results that resisted mathematicians for decades. They say their work teaches an LLM to do it, highlighting a quantitative notion of “interestingness,” 4.3× higher interestingness and a self-expanding discovery loop.
LLM reportedly taught to find interesting theorems without human guidance
A post introducing the work claims 4.3× higher “interestingness,” with a quantitative notion of that quality and a self-expanding discovery loop.
TLDR
The author argues that finding interesting theorems without human guidance is a new bottleneck as LLMs prove results that resisted mathematicians for decades. They say their work teaches an LLM to do it, highlighting a quantitative notion of “interestingness,” 4.3× higher interestingness and a self-expanding discovery loop.
