{"id":"arxiv/2610.31745","url":"https://arxiv.org/abs/2610.31745","tags":["hep-ph","hep-ex","hep-th"],"score":0,"title":"FeynFacet: AI agents for analytic NLO QCD","domain":"hep-ph","source":"arxiv","authors":["Zhong-Bo Kang","Werner Vogelsang","Congyue Zhang"],"reviews":0,"subject":"hep-ph","summary":"We present FeynFacet, a workflow in which artificial intelligence agents develop and execute programs for analytic next-to-leading-order (NLO) QCD calculations, with the observables and conventions specified by physicists. We use it to derive new analytic hard functions for single-inclusive transverse spin transfer in proton-proton and lepton-proton collisions. The new proton-proton results cover all partonic channels contributing to transverse spin transfer, and the new lepton-proton result is inclusive in the scattered lepton. Within FeynFacet, we use reverse unitarity to represent the…","category":"research","comments":0,"disputed":false,"division":"physics","mentions":0,"discussed":0,"published":"2026-10-22","reactions":{}}