{"id":"arxiv/2609.29756","url":"https://arxiv.org/abs/2609.29756","tags":["math.PR","cs.DM","cs.NE","math.CO","stat.ML"],"score":0,"title":"Boolean threshold functions, neuron capacity, and memory retrieval","domain":"math.PR","source":"arxiv","authors":["Xinyuan Xie"],"reviews":0,"subject":"math.PR","summary":"How much information can a single neuron remember? How many memories can neural networks retrieve without creating false memories? These questions are related to a basic question: how many Boolean threshold functions $f(x)=\\operatorname{sgn}(a_0+\\langle a,x\\rangle)$, $x\\in\\{-1,1\\}^n$, are there? In this paper, we show that the number $T_n$ of distinct Boolean threshold functions is \\[ T_n=2\\binom{2^n-1}{n}\\bigl(1+O(n^{-99})\\bigr). \\] Equivalently, the capacity of a single threshold neuron is $n^2-\\log_2(n!)+1+O(n^{-99})$ bits, improving the $O(n)$ error term in the result of…","category":"research","comments":0,"disputed":false,"division":"math","mentions":0,"discussed":0,"published":"2026-09-24","reactions":{}}