{"id":"arxiv/2611.13872","url":"https://arxiv.org/abs/2611.13872","tags":["math.OC"],"score":0,"title":"LongTails: Automated Research in Optimization","domain":"math.OC","source":"arxiv","authors":["Alexander Tyurin","QuocViet Pham","Valentin Khrulkov","Ivan Oseledets"],"reviews":0,"subject":"math.OC","summary":"We introduce LongTails, a framework for discovering and improving oracle-complexity bounds with formal verification in Lean 4. Using LongTails, we have already obtained more than 100 new results across a broad range of optimization problems, including generalized smoothness, heavy-tailed noise, finite-sum and composite optimization, and higher-order and zeroth-order methods. Our goal is to scale discovery and continually develop new methods with non-trivial and valuable theoretical guarantees using modern agentic systems. The framework is specialized for optimization, where we separate a…","category":"research","comments":0,"disputed":false,"division":"math","mentions":0,"discussed":0,"published":"2026-11-10","reactions":{}}