Jets with Variable R

TL;DR

The paper introduces a class of variable-$\Delta R$ jet algorithms, focusing on a concrete choice $R_{ m eff}(p_T)=\rho/p_T$ (VR) to better reflect jet sizes across boosts. Implemented within sequential recombination via AKT-VR and CA-VR, the approach is tested on resonance decays, cascade decays, and three-body gluino decays, showing typical 10–20% gains in signal efficiency over fixed-$R$ algorithms. Backgrounds are addressed with jet-quality cuts that preserve the VR gains while reducing contamination, yielding significant improvements in $S/\sqrt{B}$ (often 10–20%) in realistic scenarios. The results advocate for further exploration of designer VR variants and confirm the method’s infrared/collinear safety and boost-invariant properties, offering a practical path to enhanced jet-based analyses at hadron colliders.

Abstract

We introduce a new class of jet algorithms designed to return conical jets with a variable Delta R radius. A specific example, in which Delta R scales as 1/pT, proves particularly useful in capturing the kinematic features of a wide variety of hard scattering processes. We implement this Delta R scaling in a sequential recombination algorithm and test it by reconstructing resonance masses and kinematic endpoints. These test cases show 10-20% improvements in signal efficiency compared to fixed Delta R algorithms. We also comment on cuts useful in reducing continuum jet backgrounds.