Gauge invariance from quantum information principles

Claudia Núñez, Miguel Pardina, Manuel Asorey, José Ignacio Latorre, Alba Cervera-Lierta · 2025

Read Gauge invariance from quantum information principles by Claudia Núñez, Miguel Pardina, Manuel Asorey, José Ignacio Latorre, Alba Cervera-Lierta (2025) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2511-04358.txt.

Introduction

Claudia Núñez, Miguel Pardina, Manuel Asorey, José Ignacio Latorre, and Alba Cervera-Lierta ask whether quantum information can explain gauge invariance. Their study links two resources, entanglement and magic, to the rules that govern basic forces. Magic here means the non-Clifford part that makes universal quantum computing possible.

The authors study gluon-gluon and graviton-graviton scattering in the high-energy limit. There, mass terms fade and conformal symmetry holds, so particles act as massless. They argue that limits on entanglement and magic together restrict how such particles can interact. The work offers a fresh route from quantum information principles toward the symmetries of fundamental physics.

Catalog record

CREATORClaudia Núñez, Miguel Pardina, Manuel Asorey, José Ignacio Latorre, Alba Cervera-Lierta
YEAR2025
COLLECTIONPAPERS
FORMATTXT
FILE SIZE40 KB
VERSION1.0
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2511-04358

Browse the full library · About the Tower of Babel

Full text

The opening of the work, from the library's hosted text:

Gauge invariance from quantum information principles Claudia Núñez, Miguel Pardina, Manuel Asorey, José Ignacio Latorre, Alba Cervera-Lierta Abstract Entanglement is a hallmark of quantum theory, yet it alone does not capture the full extent of quantum complexity: some highly entangled states can still be classically simulated. Non-classical behavior also requires magic, the non-Clifford component that enables universal quantum computation. Here we investigate whether the interplay between entanglement and magic constrains the structure of fundamental interactions. We study gluon-gluon and graviton-graviton scattering at tree level, explicitly breaking gauge and general covariance by modifying the quartic vertices and analyzing the resulting generation of entanglement and magic. We find that imposing maximal entanglement (MaxEnt) alone does not uniquely recover gauge-invariant and diffeomorphism invariant interactions, but adding the condition of minimal, but nonzero, magic singles it out. Our results indicate that nature favors MaxEnt and low magic: maximal quantum correlations with limited non-Cliffordness, sufficient for universal quantum computing but close to classical simulability. This dual informational principle may underlie the emergence of gauge invariance in fundamental physics. I Introduction Entanglement is often regarded as the core feature of quantum theory, capturing correlations that cannot be explained classically 2 ; 4 . Yet, entanglement alone does not fully characterize quantum complexity: highly entangled systems can still be efficiently simulated on a classical computer 20 ; 1 . What distinguishes truly quantum behavior is the presence of magic, the non-Clifford component of a quantum state or process, that enables universal quantum computation and resists classical description. In this sense, while entanglement is necessary to go beyond local realism, magic is required to go beyond classical simulability 25 . These observations raise a fundamental question: are the interactions governing elementary particles structured so that they both generate entanglement and escape classical simulability? In other words, could the underlying principles of nature be constrained not only by symmetry requirements, such as gauge invariance, but also by quantum information–theoretical principles such as entanglement and magic generation? Recent developments support this viewpoint. By imposing maximal entanglement (MaxEnt) generation, quantum electrodynamics (QED) at tree-level can be recovered 11 , and the weak mixing angle obtained aligns closely with experimental values 11 ; 38 . Similarly, enforcing MaxEnt in gluon scattering at tree-level reproduces quantum chromodynamics (QCD) and filters almost all unphysical interactions 40 . These phenomenological works are becoming relevant as experimental progress has also made it possible to detect entanglement at the LHC 7 ; 14 ; 13 and to observe Bell inequality violations in particle processes 16 ; 17 ; 18 . On the other limit, the degree of entanglement suppression itself appears to encode information about the nature of the interactions 8 ; 35 ; 30 . Intriguingly, the generation of magic also varies among fundamental interactions. QED produces almost no magic 33 , while top quarks can generate significant magic depending on kinematics 43 . By contrast, gluon-gluon and graviton-graviton scatterings exhibit very low magic at tree level 19 . Furthermore, the weak mixing angle has been determined to \sin^{2}\theta_{W}\simeq 0.231 , very close to the experimental value, under the principle of minimal magic 31 . Collectively, these results suggest that quantum information principles, particularly those quantifying entanglement and non-Cliffordness, may play a crucial role in the structure of known interactions and in the search for physics beyond the Standard Model 5 ; 42 ; 6 ; 9 ; 15 ; 36 ; 37 ; 10 ; 27 . In this work, we explore whether gauge invariance, a cornerstone of modern field theory, can be recovered purely from quantum information principles. To this end, we study gluon-gluon and graviton-graviton scattering at tree level. As massless bosons, we take the two polarization states, left L and right R , as the degrees of freedom, therefore restricting the quantum information analysis to a two-qubit pure state. As a consequence, both entanglement and magic are well-defined monotones. We explicitly break gauge invariance in QCD and general covariance in perturbative quantum gravity by modifying the four-vertex interaction and studying how such deformations affect entanglement and magic generation. Our analysis shows that imposing MaxEnt alone does not uniquely recover the gauge-invariant interaction, but when we further impose minimal (but nonzero) magic generation, the physical gauge-invariant solution is uniquely singled out. These findings suggest that nature may indeed be guided by a dual informational principle: it favors maximal quantum correlation (MaxEnt) while maintaining low, yet nonzero, magic—enough to enable universality in quantum dynamics, but limited to preserve near-classical simulability. This interplay between entanglement and magic may thus provide a new lens through which to view the origin and structure of fundamental interactions. This work is organized as follows. First, we present how entanglement and magic can be quantified for pure states: by means of the concurrence and the Stabilizer Renyi Entropies. Next, in Sec. III we review how entanglement and magic are generated in tree-level gluon and graviton scattering. Section IV introduces the main analysis of this work: by explicitly breaking gauge-invariance, we recover the gauge-invariant result by imposing MaxEnt and minimal magic, both in gluon-gluon and graviton-graviton processes. We close the paper with some conclusions and discussion. II Entanglement and magic in two-particle processes Our ultimate goal is to study under which circumstances non-locality and non-classical simulability emerge from fundamental interactions. In this work, we restrict our analysis to two-particle processes, thereby limiting it to tree-level processes. Both gluons and gravitons are massless bosons, which impose two polarization states, |L\rangle and |R\rangle . Therefore, their quantum state can be modeled by a pure two-qubit state whose amplitudes come from the scattering amplitudes computed at tree-level: |\psi_{f}\rangle\sim\mathcal{M}_{\psi_{i}\rightarrow RR}|RR\rangle+\mathcal{M}_{\psi_{i}\rightarrow RL}|RL\rangle\\ +\mathcal{M}_{\psi_{i}\rightarrow LR}|LR\rangle+\mathcal{M}_{\psi_{i}\rightarrow LL}|LL\rangle, (1) where \psi_{i} is the initial state and \mathcal{M}_{\psi_{i}\rightarrow AB} is the tree-level amplitude for the outgoing particles having AB polarization A,B\in\{R,L\} . We dropped the ket notation for simplicity. The reduction of the problem to two-qubit pure states substantially simplifies the entanglement and non-locality analysis. Any two-qubit entangled state that is pure violates a Bell inequality 24 . Moreover, all entanglement measures are equivalent to each other for two-qubit pure states 39 . As a consequence, we can choose a convenient figure of merit to quantify the amount of entanglement generated and the result will also imply the amount of non-locality of that state. Taking a general two-qubit pure state, |\psi\rangle=\alpha|RR\rangle+\beta|RL\rangle+\gamma|LR\rangle+\delta|LL\rangle, (2) with \alpha,\beta,\gamma,\delta\in\mathbb{C} and |\alpha|^{2}+|\beta|^{2}+|\gamma|^{2}+|\delta|^{2}=1 , we quantify the entanglement using the concurrence, \Delta=2|\alpha\delta-\beta\gamma|, (3) where 0\leq\Delta\leq 1 . The states with \Delta=0 correspond to product states and the ones with \Delta=1 to MaxEnt states. Systems that generate high entanglement are not necessarily hard to simulate. Indeed, one can construct highly entangled states (even Absolutely Maximally Entangled States 23 ; 12 ) in a quantum computer using only Haddamard and Controlled-Z gates 22 and still be able to represent that state with classical resources efficiently 20 . One needs to use a quantum gate outside the Clifford group to make things hard to reproduce by a classical computer, i.e. to perform universal quantum computation. This property is typically exemplified by the number of T gates (non-Clifford gates) necessary to construct such states. In other words, the amount of non-Cliffordness in a quantum state is a measure of how hard

Read the complete text