{"id":"arxiv/2604.08859","url":"https://arxiv.org/abs/2604.08859","tags":["math.DS"],"score":0,"title":"Computing with reaction networks at input-independent speed: exponential and logarithmic functions","domain":"math.DS","source":"arxiv","authors":["David F. Anderson","Badal Joshi","Tung D. Nguyen"],"reviews":0,"subject":"math.DS","summary":"The concept of \\textit{input-independent computational speed} for chemistry-based analog computers was introduced in Anderson-Joshi (2025), where it was shown that arithmetic operations can be computed with a convergence rate bounded below independently of the input values. Here, by inputs we mean the numerical values encoded by the initial concentrations of designated input species, with the underlying reaction network and rate constants held fixed. Combining these operations via power series to approximate transcendental functions is possible in principle, but the number of chemical…","category":"research","comments":0,"disputed":false,"division":"math","mentions":0,"discussed":0,"published":"2026-04-10","reactions":{}}