Chiral higher-spin symmetry of the celestial twistor sphere
Tung Tran
TL;DR
The paper constructs and analyzes chiral higher‑spin symmetry algebras $\mathfrak{ca}$ on the celestial twistor sphere via Koszul duality from holomorphic twistorial higher‑spin theories. Classically, $\mathfrak{ca}$ is associative and identifies with color‑kinematic algebras of 4d chiral/self‑dual theories; quantum corrections can break associativity unless the spectrum is in the anomaly‑protected sets or axionic currents are added. The authors compute OPEs and a range of correlation functions, showing that tree and loop amplitudes are rational and highly constrained, with nontrivial higher‑spin form factors appearing mainly in Yang–Mills‑like setups; in many cases, loop amplitudes vanish due to quantum integrability. They also formulate defect CFTs sourcing the currents, including chiral boson/fermion models and matrix‑valued cases, and discuss extensions with Weyl fermions and axions. Overall, the work delineates when the celestial chiral algebras remain associative at one loop and what sorts of higher‑spin correlators are allowed, shedding light on possible unconstructed higher‑spin theories and their observable signatures on the celestial sphere.
Abstract
We study the chiral higher-spin symmetry algebras $\mathfrak{ca}$ of various twistorial higher-spin theories. These symmetries play the roles of asymptotic symmetries on the celestial twistor sphere, which constrain the observables of twistorial theories. To first order in quantum correction, we show that the chiral algebras associated with anomaly-free holomorphic twistorial higher-spin theories are associative themselves. On the other hand, the chiral algebras associated with anomalous holomorphic twistorial higher-spin theories only become associative upon including suitable axionic currents. When computing $4d$ form factors in terms of correlation functions between higher-spin currents on the celestial twistor sphere, we observe that there are some non-vanishing higher-spin form factors. This observation, however, is only well justified for the case of theories with Yang-Mills-like interactions. We also give some brief comments on the case of higher-derivative interactions.
