Light-Front Gluon Spectator Model

• The light-front gluon spectator model is a framework that represents a hadron as an active gluon plus an effective spectator, formulated in light-front coordinates.
• It encompasses diverse architectures such as correlator-based calculations and explicit light-front wave-function overlaps to derive gluon TMDs, PDFs, GPDs, GTMDs, and Wigner distributions.
• The model employs parameter-fitting techniques against experimental and PDF data, offering insights into gluon spin, momentum, and orbital contributions in hadron structure.

The light-front gluon spectator model denotes a class of hadron-structure constructions in which an observed gluon is treated as the active parton and the unresolved remainder of the hadron is replaced by an effective spectator system, with the dynamics formulated in light-front or light-cone variables such as $x=k^+/P^+$ and intrinsic transverse momentum. In the proton case, this usually means a two-body $g+X$ truncation with a spin-$\tfrac12$ spectator; in other applications, such as the pion, the spectator can carry different effective quantum numbers. Across the literature, the term covers both correlator-based spectator calculations and explicit light-front wave-function overlap models, all aimed at constructing gluon PDFs, TMDs, GPDs, GTMDs, Wigner distributions, gravitational form factors, and spin decompositions from a tractable nonperturbative input (Bacchetta et al., 2020, Bacchetta et al., 2021, Chakrabarti et al., 2023).

1. Conceptual definition and historical scope

In its most common proton realization, the model assumes that the parent hadron can be resolved into an active gluon plus an effective remnant that carries the complementary momentum and the remaining quantum numbers. One representative formulation states that “the proton can emit a gluon with longitudinal-momentum fraction $x$ and transverse momentum $\boldsymbol{p}_T$, and the remainders are treated as an effective colored particle with mass $M_X$ and possessing the quantum numbers of a fermion, that we call spectator” (Bacchetta et al., 2021). Closely related proton models describe the state as a two-body light-front truncation $|P,\Lambda\rangle \sim |g+\text{spectator}\rangle$, with active gluon helicity $\lambda_g=\pm 1$ and spectator helicity $\lambda_X=\pm \tfrac12$ (Chakrabarti et al., 2024, Choudhary et al., 2024).

The same logic has been applied beyond proton twist-2 TMDs. In a light-cone spectator treatment of the gluon Sivers function, the proton is approximated by an effective two-body state in which the active parton is a gluon and the spectator represents the three valence quarks grouped into a single spin-$\tfrac12$ color-carrying object (Lu et al., 2016). For gluon GPDs, GTMDs, and Wigner distributions, the spectator again plays the role of an effective $g+X$0 remnant (Tan et al., 2023, Tan et al., 2023). In mesonic applications, the same active-gluon-plus-spectator reduction appears in a pion model where the spectator is an effective system containing “the remaining constituents,” and in BLFQ treatments of heavy mesons where the $g+X$1 sector can be reinterpreted as an active gluon plus a $g+X$2 spectator subsystem (Kaur et al., 2 Jul 2025, Wu et al., 9 Mar 2026).

A useful distinction runs through the literature. Some papers use “light-front” in a strict Hamiltonian sense, with explicit Fock-state wave functions and overlap formulas; others are “light-front” in a kinematical and correlator-based sense, because they are written directly in light-cone coordinates and evaluate standard TMD correlators at fixed light-front separation without deriving the distributions from explicit bound-state wave functions. This distinction is explicit in the proton TMD spectator literature and is central to the model’s interpretation (Bacchetta et al., 2020, Bacchetta et al., 2021).

2. Light-front kinematics and model architectures

The common kinematical backbone is the light-front decomposition of momenta. In a standard spectator-TMD formulation, a four-vector is written as

$g+X$3

with

$g+X$4

The parent nucleon is often taken in a frame with no transverse momentum,

$g+X$5

and the active gluon momentum is parameterized as

$g+X$6

Imposing the spectator on-shell condition $g+X$7 then fixes the gluon virtuality in terms of $g+X$8, $g+X$9, and $\tfrac12$0 (Bacchetta et al., 2020). Twist-3 extensions use the same light-cone basis, with correlators evaluated at $\tfrac12$1 and parameterized by $\tfrac12$2, $\tfrac12$3, and transverse tensor components (Xie et al., 3 May 2026).

Within this kinematical framework, three major architectures have emerged.

Architecture	Representative papers	Distinctive ingredients
Correlator-based spectator TMD model	(Bacchetta et al., 2020, Bacchetta et al., 2021, Xie et al., 3 May 2026)	Effective vertex, tree-level correlator, continuous spectator-mass spectrum
Explicit LFWF spectator model	(Lu et al., 2016, Chakrabarti et al., 2023, Tan et al., 2023, Tan et al., 2023, Chakrabarti et al., 2024, Choudhary et al., 2024)	Two-body Fock expansion, helicity amplitudes, overlap formulas
Dynamical light-front bound-state model with explicit gluon sector	(Kaur et al., 2 Jul 2025, Wu et al., 9 Mar 2026)	Holographic/$\tfrac12$4 Hooft or BLFQ dynamics, effective spectator interpretation of higher Fock content

The correlator-based branch starts from the gauge-invariant gluon TMD correlator

$\tfrac12$5

then evaluates it in a tree-level spectator approximation (Bacchetta et al., 2020). By contrast, the explicit-LFWF branch expands the proton directly in a two-particle light-front basis with amplitudes $\tfrac12$6, and computes PDFs, TMDs, GPDs, or Wigner distributions as overlaps of those amplitudes (Chakrabarti et al., 2023, Tan et al., 2023). A more microscopic variant appears in BLFQ heavy-meson studies, where the state is solved in a truncated $\tfrac12$7 basis and the gluon-active spectator interpretation is read off from the resulting wave functions rather than imposed phenomenologically (Wu et al., 9 Mar 2026).

Perturbative light-front gluon cascade papers are relevant but conceptually distinct. They derive light-front gluon wave functions, fragmentation functions, and their relation to helicity amplitudes in LFPT, with variables such as $\tfrac12$8 and $\tfrac12$9, but they do not introduce a hadronic spectator state, a nonperturbative vertex, or PDF/TMD/GPD phenomenology (Cruz-Santiago, 2013, Cruz-Santiago et al., 2013). A plausible implication is that they supply perturbative templates rather than complete spectator models.

3. Vertices, wave functions, and spectral constructions

In the correlator-based proton TMD model, the central dynamical ansatz is the effective nucleon-gluon-spectator vertex

$x$0

The two form factors are chosen in dipolar form,

$x$1

in order to cancel the singularity of the gluon propagator, suppress large-$x$2 regions, and render $x$3-integrated quantities finite (Bacchetta et al., 2020). The proceedings formulation emphasizes the same point in less explicit form, noting that the effective vertex contains two dipolar form factors chosen as functions of $x$4 (Bacchetta et al., 2021).

A distinctive refinement is the replacement of a fixed spectator mass by a continuous spectral average,

$x$5

with a seven-parameter spectral function $x$6 combining a smooth large-$x$7 component and a Gaussian-like term (Bacchetta et al., 2020). The proceedings paper stresses that this spectral parametrization is “suited to describe both moderate and small-$x$8 effects,” and uses it precisely because a single-mass spectator model is too restrictive (Bacchetta et al., 2021). The twist-3 extension preserves the same continuous spectator-mass strategy while moving to higher-twist correlators (Xie et al., 3 May 2026).

In the explicit LFWF branch, the proton state is expanded as

$x$9

A recurring helicity structure, patterned after the dressed electron in QED, contains amplitudes proportional to

$\boldsymbol{p}_T$0

multiplying a common scalar wave function $\boldsymbol{p}_T$1 (Lu et al., 2016, Chakrabarti et al., 2023). Two popular radial choices appear repeatedly. One is a BHL-type invariant-mass regulator,

$\boldsymbol{p}_T$2

used in the Sivers, GPD, and Wigner-distribution models (Lu et al., 2016, Tan et al., 2023, Tan et al., 2023). The other is a modified soft-wall AdS/QCD wave function,

$\boldsymbol{p}_T$3

used in proton gluon PDF/TMD/GPD models and their applications to GFFs and near-threshold quarkonium production (Chakrabarti et al., 2023, Choudhary et al., 2024, Sain et al., 9 Jan 2026).

Mesonic variants modify the spectator quantum numbers and bound-state input. In the pion model, the active constituent is a massless gluon, the spectator is assigned spin 1, and the spin-independent LFWF is factorized into transverse and longitudinal parts constrained by the light-front holographic Schrödinger equation and the ’t Hooft equation (Kaur et al., 2 Jul 2025). In heavy mesons, BLFQ solves a truncated Hamiltonian in the $\boldsymbol{p}_T$4 basis, so the effective spectator description emerges dynamically rather than through a fixed phenomenological vertex (Wu et al., 9 Mar 2026).

4. Distribution functions and observable sectors

The original target of the proton spectator-TMD program was the complete set of leading-twist, $\boldsymbol{p}_T$5-even gluon TMDs accessible at tree level without gauge-link phases: $\boldsymbol{p}_T$6 In the correlator-based model these are projected from $\boldsymbol{p}_T$7 and calculated analytically at fixed $\boldsymbol{p}_T$8, then spectrally averaged (Bacchetta et al., 2020). The proceedings version emphasizes that the framework addresses “all twist-2 $\boldsymbol{p}_T$9-even gluon TMDs” and explicitly highlights the Boer–Mulders-type distribution $M_X$0, interpreted there as the density of transversely polarized gluons inside an unpolarized proton (Bacchetta et al., 2021). The AdS/QCD-inspired LFWF models compute the same $M_X$1-even set by overlap and show that the resulting TMDs satisfy positivity and Mulders–Rodrigues inequalities (Chakrabarti et al., 2023, Choudhary et al., 2024).

Beyond twist 2, the spectator approach has been extended to twist-3 gluon TMDs through the gluon-gluon correlators $M_X$2 and $M_X$3. In that extension, the relevant functions are complex, with real parts corresponding to $M_X$4-even structures and imaginary parts to $M_X$5-odd structures, and the equation-of-motion relation was checked numerically and found to hold fairly well in the spectator model (Xie et al., 3 May 2026).

The same light-front spectator logic has been used for off-forward distributions. At $M_X$6, a light-cone spectator model gives overlap representations for

$M_X$7

and predicts $M_X$8 together with the model relation

$M_X$9

(Tan et al., 2023). At nonzero skewness, an AdS/QCD-based proton spectator model computes both chiral-even and chiral-odd gluon GPDs in the DGLAP region $|P,\Lambda\rangle \sim |g+\text{spectator}\rangle$0, and reports that only six of the eight leading-twist gluon GPDs survive in that model (Chakrabarti et al., 2024). Another proton study presents chiral-even and chiral-odd GPDs in a related light-front spectator framework and uses them to infer gluon spin and OAM contributions (Choudhary et al., 2024).

The most differential phase-space observables arise in the GTMD and Wigner-distribution sector. A light-cone spectator model computes the gluon Wigner distributions $|P,\Lambda\rangle \sim |g+\text{spectator}\rangle$1, $|P,\Lambda\rangle \sim |g+\text{spectator}\rangle$2, $|P,\Lambda\rangle \sim |g+\text{spectator}\rangle$3, and $|P,\Lambda\rangle \sim |g+\text{spectator}\rangle$4, together with the associated GTMDs $|P,\Lambda\rangle \sim |g+\text{spectator}\rangle$5 and $|P,\Lambda\rangle \sim |g+\text{spectator}\rangle$6, and uses them to extract canonical gluon OAM and spin-orbit correlations (Tan et al., 2023). A dressed-quark light-front model, while not a proton spectator model in the phenomenological sense, provides an explicit two-body benchmark for the same gluon Wigner and GTMD structures and also distinguishes canonical and kinetic gluon OAM (Mukherjee et al., 2015).

The observable range extends further once GPD moments are taken. In a proton light-front gluon-spectator model inspired by soft-wall AdS/QCD, matrix elements of the gluonic energy-momentum tensor are used to obtain the gluon gravitational form factors $|P,\Lambda\rangle \sim |g+\text{spectator}\rangle$7, $|P,\Lambda\rangle \sim |g+\text{spectator}\rangle$8, $|P,\Lambda\rangle \sim |g+\text{spectator}\rangle$9, and $\lambda_g=\pm 1$0, which are then inserted into near-threshold $\lambda_g=\pm 1$1 and $\lambda_g=\pm 1$2 photoproduction amplitudes (Sain et al., 9 Jan 2026). In the pion, the same active-gluon spectator logic has been used to compute gluon PDFs, GPDs, TMDs, impact-parameter distributions, longitudinal diffraction patterns, and the pion gravitational form factor (Kaur et al., 2 Jul 2025). In heavy mesons, BLFQ with an explicit $\lambda_g=\pm 1$3 sector yields the first gluon PDFs in that framework, together with electromagnetic form factors, decay constants, and PDAs (Wu et al., 9 Mar 2026).

5. Parameter determination and phenomenological outputs

Parameter fixing is model dependent and constitutes a defining part of the spectator approach. In the correlator-based proton TMD model, the full parameter set contains ten free parameters,

$\lambda_g=\pm 1$4

with $\lambda_g=\pm 1$5 fixed to $\lambda_g=\pm 1$6 because the fit is insensitive to it (Bacchetta et al., 2020). The parameters are constrained by matching

$\lambda_g=\pm 1$7

to NNPDF3.1sx and NNPDFpol1.1 at

$\lambda_g=\pm 1$8

over

$\lambda_g=\pm 1$9

using a bootstrap procedure with $\lambda_X=\pm \tfrac12$0 replicas (Bacchetta et al., 2020). The proceedings summary stresses the same bootstrap logic, namely repeated Gaussian-noise replicas of the NNPDF input fitted separately to propagate PDF uncertainties into the spectator-model parameter space (Bacchetta et al., 2021).

In the AdS/QCD-inspired proton LFWF model for gluon PDFs and TMDs, four parameters are used,

$\lambda_X=\pm \tfrac12$1

and are fitted to the NNPDF3.0 NLO unpolarized gluon PDF at $\lambda_X=\pm \tfrac12$2 GeV using 300 data points in $\lambda_X=\pm \tfrac12$3 and 100 replicas (Chakrabarti et al., 2023). The paper reports

$\lambda_X=\pm \tfrac12$4

with

$\lambda_X=\pm \tfrac12$5

(Chakrabarti et al., 2023). The earlier Sivers-function model instead fixes $\lambda_X=\pm \tfrac12$6 by fitting the gluon PDF to GRV98 LO and NLO over $\lambda_X=\pm \tfrac12$7, and selects $\lambda_X=\pm \tfrac12$8 for the LO fit and $\lambda_X=\pm \tfrac12$9 for the NLO fit (Lu et al., 2016).

Illustrative phenomenological outputs are correspondingly diverse. In the correlator-based TMD model, $\tfrac12$0 is positive and clearly non-Gaussian in $\tfrac12$1, with a relatively flat tail up to $\tfrac12$2, while $\tfrac12$3 starts from a finite nonzero value at $\tfrac12$4 and decreases rapidly (Bacchetta et al., 2020, Bacchetta et al., 2021). The same model gives

$\tfrac12$5

at $\tfrac12$6 GeV (Bacchetta et al., 2020). In the AdS/QCD proton model, the average gluon longitudinal momentum fraction is

$\tfrac12$7

and the integrated gluon helicity contributions are quoted as

$\tfrac12$8

$\tfrac12$9

$g+X$00

(Chakrabarti et al., 2023).

Angular-momentum results vary strongly across models and scales. At $g+X$01, one light-cone spectator model for gluon GPDs finds

$g+X$02

(Tan et al., 2023). A nonzero-skewness proton GPD/GTMD model reports

$g+X$03

(Chakrabarti et al., 2024). The Wigner-distribution model obtains a total canonical gluon OAM of approximately

$g+X$04

and negative gluon spin-orbit correlation (Tan et al., 2023). The near-threshold photoproduction application gives

$g+X$05

and reports good agreement of the resulting $g+X$06 near-threshold photoproduction cross sections with recent Jefferson Lab data as well as earlier SLAC and Cornell measurements (Sain et al., 9 Jan 2026). These differences are a direct indication of model dependence rather than an internal inconsistency of the framework.

6. Theoretical status, limitations, and extensions

A recurring misconception is that every light-front gluon spectator model is a full light-front wave-function theory derived from QCD. The literature explicitly rejects that identification in many cases. The correlator-based TMD models are “light-front” because they use light-cone coordinates, fixed-$g+X$07 correlators, and the standard variables $g+X$08, but they are not Hamiltonian light-front constituent models with explicit Fock-state overlap formulas (Bacchetta et al., 2020, Bacchetta et al., 2021). Conversely, the LFWF-based proton, pion, and Wigner-distribution models are explicit overlap constructions but still rely on phenomenological wave-function ansätze or truncated spectator sectors rather than first-principles QCD dynamics (Chakrabarti et al., 2023, Kaur et al., 2 Jul 2025, Tan et al., 2023).

Gauge-link dependence is the central unresolved issue in much of the proton gluon-spectator literature. The correlator-based proton TMD program notes the distinction between Weizsäcker–Williams and dipole gauge-link structures, but at tree level neglects gauge-link effects and process dependence, so only $g+X$09-even leading-twist gluon TMDs are obtained (Bacchetta et al., 2020). The proceedings update states explicitly that the model “does not incorporate any gauge-link dependence” and that extension to twist-2 $g+X$10-odd gluon TMDs is underway (Bacchetta et al., 2021). By contrast, the Sivers-function model introduces a future-pointing Wilson line appropriate to SIDIS and mimics the required phase by an interaction kernel,

$g+X$11

thereby generating a $g+X$12-odd gluon Sivers function in the valence-$g+X$13 region (Lu et al., 2016). A plausible implication is that truly process-dependent gluon TMD phenomenology requires an explicit treatment of Wilson-line dynamics rather than a pure tree-level spectator correlator.

Other limitations recur across subfields. Several proton GPD models are restricted to the DGLAP region $g+X$14, because ERBL support would require particle-number-changing overlaps beyond the retained Fock sector (Chakrabarti et al., 2024). Many models are presented only at an initial hadronic scale and do not implement explicit TMD evolution (Bacchetta et al., 2020, Chakrabarti et al., 2023, Choudhary et al., 2024). Small-$g+X$15 behavior is often mimicked through spectral functions or modified endpoint factors rather than generated by BFKL or JIMWLK dynamics; the proceedings paper explicitly identifies BFKL resummation as a future direction (Bacchetta et al., 2021). In the pion and heavy-meson sectors, the effective spectator is a collective degree of freedom whose microscopic color and composition are not resolved (Kaur et al., 2 Jul 2025, Wu et al., 9 Mar 2026).

Despite these caveats, the framework has become a flexible laboratory for explicit gluonic structure. It supports proton, pion, and heavy-meson applications; it connects PDFs, TMDs, GPDs, GTMDs, Wigner distributions, GFFs, and spin decompositions through a common light-front language; and it provides a spectrum of model complexity ranging from tree-level spectator correlators to bound-state Hamiltonian calculations with explicit gluonic Fock sectors (Choudhary et al., 2024, Kaur et al., 2 Jul 2025, Wu et al., 9 Mar 2026). The most precise characterization is therefore not a single model but a family of light-front gluon-active spectator constructions whose common core is the reduction

$g+X$16

implemented with varying degrees of dynamical input and phenomenological constraint.