Critical Majorana fermion at a topological quantum Hall bilayer transition Cristian Voinea, Wei Zhu, Nicolas Regnault, Zlatko Papić Abstract Quantum Hall bilayers are a uniquely tunable platform that can realize continuous transitions between distinct topological phases of matter. One prominent example is the transition between the Halperin state and the Moore–Read Pfaffian, long predicted to host a critical theory of Majorana fermions but so far not verified in unbiased microscopic simulations. Using the fuzzy sphere regularization, we identify the low-energy spectrum at this transition with the 3D gauged Majorana conformal field theory. We show that the transition is driven by the closing of the neutral fermion gap, and we directly extract the operator content in both integer and half-integer spin sectors. Our results resolve the long-standing question of the nature of a topological phase transition in a setting relevant to quantum Hall experiments, while also providing the first realization of a fermionic theory on the fuzzy sphere, previously limited to bosonic theories. Introduction.— The fractional quantum Hall (FQH) effect has long served as a fertile ground for discovering new phases of matter that defy the Landau paradigm Tsui et al. 1982 ; Laughlin 1983 . It brought into the spotlight many-body phases with exotic quasiparticles called anyons Leinaas and Myrheim 1977 ; Wilczek 1982 , some of which could prove useful for fault-tolerant quantum computing Nayak et al. 2008 ; Pachos 2012 . A prominent example is the even-denominator state at filling factor \nu=5/2 Willett et al. 1987 , believed to be described by the Moore–Read wave function Moore and Read 1991 . The elementary excitations of the Moore–Read state are anyons with charge e/4 and non-Abelian braiding statistics Nayak and Wilczek 1996 ; Read and Green 2000 . Due to the existence of anyons, it is generally challenging to study topological phase transitions between different FQH states. For example, despite theoretical evidence Morf 1998 ; Rezayi and Haldane 2000 , experiments have yet to unambiguously determine the nature of the \nu=5/2 state Banerjee et al. 2018 ; Zibrov et al. 2017 ; Li et al. 2017 ; Zibrov et al. 2018 ; Falson et al. 2018 ; Dutta et al. 2021 ; Huang et al. 2022 ; Liu et al. 2022 ; Farahi et al. 2023 ; Willett et al. 2023 ; Hu et al. 2025a . Multicomponent quantum Hall systems, such as bilayers and wide quantum wells, offer a complementary platform for exploring a broader class of FQH states and transitions between them Girvin and MacDonald 2007 ; Eisenstein 2014 . These systems are highly tunable, as the effective interactions and tunneling can be controlled via electrostatic gates and by varying the magnetic length, see Fig. 1 (a). Such tuning can induce quantum phase transitions between FQH phases, including the one at filling \nu=1/2 , observed in numerous experiments Suen et al. 1992 ; Eisenstein et al. 1992 ; Suen et al. 1994 ; Shabani et al. 2013 The transition is believed to occur between a two-component Halperin state Halperin 1983 and a single-component Moore–Read state He et al. 1993 ; Nomura and Yoshioka 2004 ; Papić et al. 2010 ; Peterson and Das Sarma 2010 ; Peterson et al. 2010 ; Liu et al. 2016 ; Zhu et al. 2016 ; Cabra et al. 2001 ; Cappelli et al. 2001 ; Repellin et al. 2015 ; Crépel et al. 2019a , Fig. 1 (b). Both of these gapped phases represent weakly-paired states of composite fermions Jain 1989 . The tunneling-driven critical point between them is predicted to be governed by a massless Majorana fermion Wen 2000 ; Read and Green 2000 ; Barkeshli and Wen 2010 ; Barkeshli and Wen 2011 – the long-sought particle that is its own antiparticle Elliott and Franz 2015 ; Yazdani et al. 2023 . While the \nu=1/2 transition was analyzed at the level of mean-field theory in Ref. Read and Green 2000 more than two decades ago, its identification in a microscopic model has remained elusive. Figure 1: (a) Quantum Hall bilayer system with the layers \uparrow , \downarrow . The particles interact via intralayer ( V_{\mathrm{intra}} ) and interlayer ( V_{\mathrm{inter}} ) interactions, and they can tunnel between the layers with an amplitude h . (b) At a critical tunneling h_{c} , the bosonic particles at filling \nu=1 undergo a transition between the 220 Halperin state and the Moore–Read state. (c) At criticality, the energy spectrum on the fuzzy sphere, resolved as a function of angular momentum, matches that of the free 3D Majorana fermion conformal field theory, with the characteristic towers of primary fields (labelled) and their descendants. In this paper, we employ the recently developed fuzzy-sphere regularization Zhu et al. 2023 to provide the first microscopic verification of the 3D Majorana conformal field theory (CFT) emerging in the quantum Hall bilayer in the presence of interlayer tunneling. To minimize finite-size effects, we assume bosonic particles at filling \nu=1 and study the transition between the two-component 220 Halperin (i.e., akin to two decoupled Laughlin states at \nu=1/2 ) and a single-component Moore–Read state. We numerically demonstrate the closing of the gap and extract the operator content of the CFT, including both integer and half-integer spin sectors, as summarized in Fig. 1 (c). These results establish fingerprints of an emergent Majorana fermion in a setting applicable to quantum Hall experiments. Beyond quantum Hall, they also open a new direction for the fuzzy sphere studies of 3D CFTs, which so far have been restricted to bosonic fields. Gauged Majorana field theory.— The continuous Halperin-Pfaffian transition is driven by charge-neutral pairs of quasiparticles and quasiholes shared between the two layers Read and Green 2000 ; Wen 2000 . The critical theory is defined by the Lagrangian \mathcal{L}=\bar{\psi}(i\gamma_{\mu}\partial^{\mu}-m)\psi\,, (1) in which the Majorana fermion \psi appears as a 2-component spinor that is indistinguishable from its antiparticle. The massless m=0 case enjoys an enlarged conformal symmetry. This symmetry imposes strong constraints on the operator content, organizing operators in conformal towers corresponding to different primaries of the theory Francesco et al. 1997 ; Zhu et al. 2023 , as summarized in Fig. 1 (c). The non-interacting Majorana field brings additional constraints. On the one hand, the equation of motion, \gamma_{\mu}\partial^{\mu}\psi=0 , fixes the descendants of \psi to only \partial_{\mu_{1}}\dots\partial_{\mu_{l}}\psi , with scaling dimension \Delta=l+1 and SO(3) angular momentum L=l+1/2 . On the other hand, the reality condition also reduces the number of operators. For example, this theory does not have a conserved vector current, as J^{\mu}=\bar{\psi}\gamma^{\mu}\psi vanishes identically. Through Fierz identities, this also implies that any simple multiple-fermion terms beyond \psi and \bar{\psi}\psi vanish. The emergence of conformal symmetry can be observed in microscopic models defined on spherical manifolds S^{d-1}\times\mathbb{R} through the state-operator correspondence Cardy 1984 ; Cardy 1985 . While historically this has presented difficulties for lattice models in d>2 , the standard setup of FQH spherical geometry with a magnetic monopole placed at the center of the sphere Haldane 1983 elegantly removes these problems. This new “fuzzy sphere” method of regularization Zhu et al. 2023 has already proven successful at extracting spectral data of many bosonic 3D CFTs Zhu et al. 2023 ; Hu et al. 2023 ; Hu et al. 2025b ; Han et al. 2023 ; Hofmann et al. 2024 ; Zhou and Zou 2025 ; Fardelli et al. 2025 ; Han et al. 2024 ; Fan et al. 2025 ; Cruz et al. 2025 ; Miro and Delouche 2025 ; Zhou and He 2025 ; Zhou et al. 2024 ; Yang et al. 2025 ; Zhou et al. 2025 ; Taylor et al. 2025 ; He 2025 . These studies realized CFT critical points by tuning an integer quantum Hall state of electrons through a spin transition. However, as shown in Ref. Voinea et al. 2025 , the fuzzy sphere regularization remains powerful at fractional fillings, which also allows to use bosons as underlying degrees of freedom. We will follow the latter approach to directly extract the universal description of the Halperin-to-Pfaffian transition from the fuzzy sphere numerics. We note that the Halperin-Pfaffian critical point differs in a few subtle respects from the conventional 3D Majorana CFT. The gauging of the global \mathbb{Z}_{2} layer-exchange symmetry Barkeshli and Wen 2011 ; Teo et al. 2015 constrains the excitations of half-integer spin, such as \psi , to live in a different Hilbert space compared to the integer-spin ones, such as \bar{\psi}\psi and T^{\mu\nu} . This is equivalent to the insertion of a gauge charge at the center of the sphere; specifically, increasing both the magnetic flux and the number of particles by 1 , the insertion of a 2\pi -flux of U(1) electric charge is bounded with a \mathbb{Z}_{2} gauge charge. For this reason, the elementary field \psi is a non-local fermion that carries gauge charge, and our critical theory is the gauged 3D Majorana fermion. In our numerics below, we will identify the Hilbert space sectors and focus on the bare fermion field \psi , the fermion mass term \bar{\psi}\psi , the stress energy tensor T^{\mu\nu}=\bar{\psi}\gamma^{\mu}\partial^{\nu}\psi , and also the next half-integer spin primary \bar{\psi}\psi\partial^{\mu}\psi , see Supplementary Material (SM) SM for further details. Model and phase diagram.— We consider bosons at Landau level filling \nu=1 on a fuzzy sphere. The monopole strength 2Q is related to the number of particles N as 2Q=\nu^{-1}N-\mathcal{S} , where \mathcal{S}=2 is the Wen-Zee shift Wen and Zee 1992 . As illustrated in Fig. 1 (a), the Hamiltonian H=H_{\text{intra}}+H_{\text{inter}}+H_{\text{t}} contains intra- and interlayer interactions, as well as the tunneling term. Using the bosonic operator b_{a}^{\dagger}(\mathbf{\Omega}) —which creates a particle in layer a=\uparrow,\downarrow at a spherical angle \mathbf{\Omega} —and the corresponding density operator, n_{a}(\mathbf{\Omega})\equiv b_{a}^{\dagger}(\mathbf{\Omega})b_{a}(\mathbf{\Omega}) , the three terms in the Hamiltonian are given by \displaystyle H_{\text{intra}} \displaystyle=\sum_{a=\uparrow,\downarrow}\int V^{\text{intra}}(\mathbf{\Omega}_{12})\,:n_{a}(\mathbf{\Omega}_{1})n_{a}(\mathbf{\Omega}_{2}):\mathrm{d}^{2}\mathbf{\Omega}_{1}\mathrm{d}^{2}\mathbf{\Omega}_{2}\,, \displaystyle H_{\text{inter}} \displaystyle=2\int V^{\text{inter}}(\mathbf{\Omega}_{12})\,n_{\uparrow}(\mathbf{\Omega}_{1})n_{\downarrow}(\mathbf{\Omega}_{2})\,\mathrm{d}^{2}\mathbf{\Omega}_{1}\mathrm{d}^{2}\mathbf{\Omega}_{2}\,, (2) \displaystyle H_{\text{t}} \displaystyle=-h\int\left(b_{\uparrow}^{\dagger}(\mathbf{\Omega})b_{\downarrow}(\mathbf{\Omega})+\mathrm{h{.}c{.}}\right)\mathrm{d}^{2}\mathbf{\Omega}\,. Here, h is the tunneling amplitude, and we take the interactions to be short-ranged, parametrized exclusively by V_{0}^{\text{intra}} (set to 1 ) and V_{0}^{\text{inter}} Haldane pseudopotentials Haldane 1983 . The model above is invariant under SO(3) rotations, while finite h leaves an additional \mathbb{Z}_{2} symmetry that exchanges the two layers. At weak tunneling h\ll V^{\text{intra}},V^{\text{inter}} , the ground state is the 220 Halperin state Halperin 1983 : \Psi_{220}(\{u_{i}^{a},v_{i}^{a}\})=\prod_{i0 in Eq. ( Critical Majorana fermion at a topological quantum Hall bilayer transition ), the gapless fermion lies in the \mathbb{Z}_{2} -odd sector. Consequently, to identify the operator spectrum of the 3D Majorana CFT we focus on two different sectors: the odd-particle, \mathbb{Z}_{2} -odd sector, corresponding to fermionic operators, and the even-particle, \mathbb{Z}_{2} -even sector, containing bosonic operators. Figure 3: (a) The cost function \delta_{\mathrm{tow}} quantifying the conformal structure in the low-lying spectrum across the phase diagram at a fixed system size N=12 . A sharp peak occurs along the dashed line at V_{0}^{\text{inter}}=0.48 . The optimal point (white cross) is further determined by the gap vanishing in (b). (b) Extrapolated gaps of \{\bar{\psi}\psi,\partial\bar{\psi}\psi,T\} at the optimal point (V_{0}^{\text{inter}},h)=(0.48,0.58) , identified from the minimum of the gaplessness cost function \delta_{\text{gap}} shown in the inset. Although not explicitly enforced, the gap of the neutral fermion \psi also converges to zero at this point, a signature of a quantum critical point in both even- and odd-particle sectors. Before analyzing the spectrum, we need to identify the optimal point with minimal finite-size effects along the phase boundary in Fig. 2 (a). We use two complementary approaches for this. First, to narrow down the space of parameters, we perform an optimization over the spectrum’s conformal structure. We consider states in the (fixed) N -even sector with \Delta\leq 4 and L\leq 2 , with the conformal tower cost function \delta_{\text{tow}}=|\bm{\Delta}|^{2}-|\bm{\Delta}\cdot\mathbf{E}|^{2}/|\mathbf{E}|^{2} , where \bm{\Delta} is the vector containing the operators’ CFT scaling dimensions and \mathbf{E} is the vector of the corresponding microscopic energies. We plot \delta_{\text{tow}} across the phase diagram in Fig. 3 (a), revealing a strong peak in the vicinity of V_{0}^{\text{inter}}\approx 0.48 . We fix this value of V_{0}^{\text{inter}} as optimal, and study the finite-size scaling of the excitation gap to determine the optimal h . Based on CFT prediction, \Delta E{\sim}1/R{\sim}1/\sqrt{2Q+1} , we construct the cost function \delta_{\text{gap}}=\sum_{i}\Delta E_{i}(N{\to}\infty) that minimizes the sum of linear 1/R extrapolations for a set of gaps corresponding to \{\bar{\psi}\psi,\partial^{\mu}(\bar{\psi}\psi),T^{\mu\nu}\} . The minimum of \delta_{\text{gap}} is achieved at the optimal h=0.58 , for which the gaps of previously mentioned states vanish, Fig. 3 (b). The (slightly) negative values of these extrapolated gaps in Fig. 3 (b) are attributed to the remnant finite-size effects which, as expected, are more pronounced for higher-energy states. As a consistency check, we also computed the lowest gap in the N -odd sector, corresponding to the neutral fermion \psi . Since this sector contains no conformal vacuum, the gap is calculated with respect to the averaged vacuum energy of the adjacent even system sizes ( N\pm 1 ), and the convergence to zero is remarkably accurate. Figure 4: Comparison between CFT data and the Hamiltonian spectrum at the optimal point (V_{0}^{\text{inter}},h)=(0.48,0.58) . (a) Even-particle sector for N=6,8,10,12,14 , containing integer angular momentum states. The lowest energy state in L=2 sector is taken to be the stress-energy tensor, whose energy was fixed to \Delta=3 . (b) Odd-particle sector for N=7,9,11,13 , containing half-integer angular momentum states. Energy levels are normalized with respect to the mean energies of the vacuum and stress-energy tensors in the adjacent, N\pm 1 , even-particle sectors. Both panels show the complete energy spectra for levels with \Delta\leq 4.5 , L\leq 7/2 . The expected CFT operators are labeled, with their scaling dimensions shown by dashed lines. Other signatures of the free Majorana CFT in these spectra also include the absence of certain states, such as the conserved vector current, J^{\mu}=\bar{\psi}\gamma^{\mu}\psi , at (L=1,\Delta=2) , and the state at (L=1/2,\Delta=3) , which is absent due to the equation of motion \gamma_{\mu}\partial^{\mu}\psi=0 . Conformal spectrum.— State-operator correspondence at the optimal critical point is shown in Fig. 4 . The ( N -even, \mathbb{Z}_{2} -even) sector can be seen in Fig. 4 (a), where all energies have been rescaled such that the stress-energy tensor—the lowest-lying state at L=2 —has \Delta=3 Zhu et al. 2023 . This sector is identified with the integer angular momentum sector of the 3D Majorana CFT, and a good agreement is obtained for the primaries \bar{\psi}\psi , T^{\mu\nu} , and their descendants. The absence of a conserved vector current, J^{\mu}=\bar{\psi}\gamma^{\mu}\psi at (L=1,\Delta=2) is also notable, further supporting the Majorana CFT nature of our critical point. Similarly, Fig. 4 (b) shows the ( N -odd, \mathbb{Z}_{2} -odd) sector, which we identify as the half-integer angular momentum sector of the CFT. This sector contains no conformal vacuum, hence we rescale each system size using the average energies of the vacuum and stress-energy tensor of the adjacent even sizes ( N\pm 1 ). State-operator correspondence is clearly manifested for the primaries \psi and \bar{\psi}\psi\partial^{\mu}\psi and a few of their descendants. The non-interacting nature of the critical point is also highlighted by the absence of the \square\psi descendant at (L=1/2,\Delta=3) . A few comments are in order. For higher levels, \Delta\geq 4 , the state-operator correspondence becomes less clear, with strong finite-size effects and additional states mixing in with the conformal towers. In other theories such as the 3D Ising CFT, conformal perturbation theory was useful in quantifying such finite-size effects Lao and Rychkov 2023 ; Läuchli et al. 2025 . However, the sparse operator-product expansion structure of the free Majorana CFT implies that the fermion mass \bar{\psi}\psi —the only relevant scalar operator—does not produce corrections to the lowest excited states in either even- or odd-particle sectors. Hence, the conformal perturbation approach cannot be directly applied to improve the matching of microscopic spectra with CFT beyond Fig. 4 . Furthermore, as in previous work Voinea et al. 2025 , our model realizes a CFT which is space-time parity symmetric, but with no corresponding symmetry at the microscopic level. A potential operator that implements the parity symmetry may be related to the emergent particle-hole (PH) symmetry of spinful bosons at \nu=1 Wang and Senthil 2016 ; Geraedts et al. 2017 . However, the microscopic parity operator appears non-trivial to write down since the our calculations are performed at a non-PH-symmetric flux value, dictated by the Wen-Zee shift of the Pfaffian and Halperin states. Experimental implications.— Droplets of bosonic FQH states have recently been realized in ultracold atoms Léonard et al. 2023 and cavity-QED experiments Wang et al. 2024 . While the ultra short-range V_{0} interactions in our model are native to such platforms, the two-component and Moore–Read states have yet to be realized. Moreover, such platforms are currently limited to a few bosons and far from the scaling regime required to observe the critical behavior. In solid state materials, the fermionic version of the transition studied above occurs at filling \nu=1/2 between the 331 Halperin and (fermionic) Moore–Read state. There is extensive experimental evidence for the 331 Halperin state and transition in both GaAs bilayers and wide quantum wells Suen et al. 1992 ; Eisenstein et al. 1992 ; Suen et al. 1994 ; Shabani et al. 2013 . However, the nature of the one-component state past the transition has remained unclear. Furthermore, these studies were limited to transport, which probes the charge gap of the system and the latter may remain open throughout the transition Zhu et al. 2023 . By contrast, light-scattering probes, as in the recent experiment Liang et al. 2024 , directly probe neutral excitations and would detect the transition. In a bilayer heterostructure with independent electric contacts to each layer Spielman et al. 2000 , it is possible to directly measure the interlayer tunneling current. The temperature dependence of this current is predicted to be \langle\varphi^{\dagger}_{\uparrow}\varphi_{\downarrow}(x)\rangle\sim T^{2} , since the tunneling term effectively relates to the mass of Majorana fermions, \bar{\psi}\psi(x) . Verification of this scaling would not only serve as evidence for the Majorana nature of the critical point, but it would also lend strong support to the identification of the Moore–Read state in the gapped phase. Conclusions.— We have microscopically established the emergence of a gauged 3D Majorana fermion at the Halperin–Pfaffian transition. Using the fuzzy sphere regularization, we have identified the hallmarks of the corresponding CFT: gap closing and operator content. These results not only offer a fresh perspective on the long-standing problem of non-Abelian states in quantum Hall bilayers, but also provide the first unbiased microscopic dissection of a continuous FQH transition, which could be directly applied to other universality classes Barkeshli and Wen 2011 ; Zhang et al. 2023 ; Barkeshli and McGreevy 2014 . It would also be important to study interfaces between Halperin and Pfaffian states Yang 2017 ; Ma and Yang 2022 ; Crépel et al. 2019b , which are directly tied to the critical theory discussed here. Beyond the quantum Hall setting, our work extends the fuzzy sphere framework to fermionic CFTs, setting the stage for applications to interacting theories such as Yukawa CFTs Fei et al. 2016 ; Gao et al. 2025 . Note Added: During the completion of this work, we became aware of Ref. Zho that realized the free Majorana fermion CFT in a different microscopic model. Acknowledgements. Acknowledgements.— We thank Zheng Zhou and Yin-Chen He for useful discussions, and for sharing their unpublished results with us. C.V. and Z.P. acknowledge support by the Leverhulme Trust Research Leadership Award RL-2019-015 and EPSRC Grant EP/Z533634/1. Statement of compliance with EPSRC policy framework on research data: This publication is theoretical work that does not require supporting research data. This research was supported in part by grant NSF PHY-2309135 to the Kavli Institute for Theoretical Physics (KITP). Computational portions of this research have made use of DiagHam dia and FuzzifiED Zhou 2025 software libraries, and they were carried out on Aire, part of the High-Performance Computing facilities at the University of Leeds. W.Z. was supported by NSFC under No. 12474144. The Flatiron Institute is a division of the Simons Foundation. Supplemental Online Material for “Critical Majorana fermion at a topological quantum Hall bilayer transition” Cristian Voinea 1 , Wei Zhu 2,3 , Nicolas Regnault 4,5,6 , and Zlatko Papić 1 1 School of Physics and Astronomy, University of Leeds, Leeds LS2 9JT, United Kingdom 2 Institute of Natural Sciences, Westlake Institute for Advanced Study, Hangzhou 310024, China 3 Department of Physics, School of Science, Westlake University, Hangzhou 310030, China 4 Center for Computational Quantum Physics, Flatiron Institute, 162 5th Avenue, New York, NY 10010, USA 5 Department of Physics, Princeton University, Princeton, New Jersey 08544, USA 6 Laboratoire de Physique de l’Ecole normale supérieure, ENS, Université PSL, CNRS, Sorbonne Université, Université Paris-Diderot, Sorbonne Paris Cité, 75005 Paris, France In this Supplementary Material, we provide (i) a more detailed introduction to the gauged 3D Majorana CFT, and (ii) an extended analysis of the different types of neutral and charge gaps at the critical point. SI S1. Gauged Majorana fermion CFT The Lagrangian for the free Majorana fermion in 3D takes the form \mathcal{L}=\bar{\psi}(i\gamma^{\mu}\partial_{\mu}-m)\psi\,, (S1) where \psi=(\psi_{1},\psi_{2})^{\mathrm{T}} is a two-component Majorana fermion field with \psi_{1,2} Grassmann fields, and \bar{\psi}=\psi^{\mathrm{T}}(i\gamma^{0}) . Using the Lorentzian signature (-1,1,1) , the chosen \gamma matrix representation \gamma^{0}=i\sigma^{2}\,,\quad\gamma^{1}=\sigma^{1}\,,\quad\gamma^{2}=\sigma^{3}\,, (S2) obeys the known Clifford algebra \{\gamma^{\mu},\gamma^{\nu}\}=2\eta^{\mu\nu} . Let us list the low-lying operator content of the theory. The fermion operator \psi has angular momentum L=1/2 and scaling dimension \Delta_{\psi}=1 . Its two-point correlation function is \langle\psi^{\alpha}(x_{1})\psi_{\beta}(x_{2})\rangle=\frac{ic_{\psi}(x_{12})^{\alpha}_{\ \beta}}{|x_{12}|^{2\Delta_{\psi}+1}}\,, (S3) where we use the notation x^{\alpha}_{\ \beta}=x^{\mu}(\gamma_{\mu})^{\alpha}_{\ \beta} and the constant c_{\psi} is fixed (see below). The equation of motion, \gamma_{\mu}\partial^{\mu}\psi=0 , imposes strong constraints on the conformal tower of states of the fermion \psi , such that the only allowed descendants are \partial_{\mu_{1}}\dots\partial_{\mu_{l}}\psi,\quad\text{with scaling dimension}\;\Delta=l+1\quad\text{and SO(3) angular momentum}\;L=l+1/2. (S4) The fermion bilinear \bar{\psi}\psi , which is space-time parity odd, is the only relevant scalar with \Delta_{\bar{\psi}\psi}=2 and L=0 . Its descendant family can be similarly constructed as \partial_{\mu_{1}}\dots\partial_{\mu_{l}}\square^{n}\bar{\psi}\psi,\quad\text{with}\quad\Delta=l+2n+2,\;L=l\,. (S5) Note that, in order to properly normalize the two-point function of the singlet to \langle\bar{\psi}\psi(x)\bar{\psi}\psi(0)\rangle=|x|^{-2\Delta_{\bar{\psi}\psi}} , we set the normalization constant to c_{\psi}=1/2 . Due to the charge-neutrality of the Majorana, the charge current J^{\mu}=\bar{\psi}\gamma^{\mu}\psi vanishes identically; moreover, with no chiral symmetry in odd dimensions, this theory does not possess any conserved currents at L=1 . Therefore, the stress-energy tensor is the next spinning operator with L=2 and \Delta_{T}=3 . For general spinning operators, we can construct two types of descendant families. The first parity-preserving family is given by the operators \partial_{\nu_{1}}\dots\partial_{\nu_{j}}\partial_{\mu_{1}}\dots\partial_{\mu_{i}}\square^{n}O_{\mu_{1}\dots\mu_{l}}\quad\text{with}\quad\Delta_{O}+2n+i+j,\;\;\;L=l+j-i. (S6) The second family, \epsilon_{\mu_{k}\rho\tau}\partial_{\rho}\partial_{\nu_{1}}\dots\partial_{\nu_{j}}\partial_{\mu_{1}}\dots\partial_{\mu_{i}}\square^{n}O_{\mu_{1}\dots\mu_{l}},\quad\text{with}\quad\Delta_{O}+2n+i+j+1,\;\;\;L=l+j-i\;, (S7) reverses the space-time parity of O . In the case of the stress energy tensor, the additional conservation law \partial_{\mu}T_{\mu\nu}=0 restricts the above families of descendants to i=0 . Moving to higher multiple-fermion operators, many conformal families vanish altogether due to the Majorana nature of the fermion. Since there are only two independent components \psi_{1},\psi_{2} , operators such as (\bar{\psi}\psi)^{2} or \psi(\bar{\psi}\psi) vanish identically. However, the field derivative \partial\psi provides independent spinor fields, with which additional primaries can be constructed. The lowest-lying operator with such a form is (\bar{\psi}\psi)\partial^{\mu}\psi , with angular momentum L=3/2 and \Delta_{(\bar{\psi}\psi)\partial\psi}=4 . Table S1 lists all operators belonging to the conformal families above with scaling dimension \Delta\leq 4 and angular momentum L\leq 3 . The exact data is compared to the fuzzy sphere data for the model in the main text at system size N=14 , rescaled such that the stress-energy tensor has \Delta_{T}=3 Zhu et al. 2023 . spin ( L ) layer \mathbb{Z}_{2} space-time parity Operator \Delta_{\text{theory}} \Delta_{\text{FS}} ( N=14 ) 1/2 -1 1 \psi 1.000 0.886 3/2 -1 1 \partial_{\mu}\psi 2.000 2.162 1/2 -1 1 \square\psi=0 NA - 5/2 -1 1 \partial_{\mu}\partial_{\nu}\psi 3.000 3.426 7/2 -1 1 \partial_{\mu}\partial_{\nu}\partial_{\rho}\psi 4.000 3.536 3/2 -1 1 \square\partial_{\mu}\psi=0 NA - 3/2 -1 -1 \bar{\psi}\psi\partial_{\mu}\psi 4.000 3.921 0 1 -1 \bar{\psi}\psi 2.000 1.862 1 1 -1 \partial_{\mu}(\bar{\psi}\psi) 3.000 3.015 2 1 -1 \partial_{\mu}\partial_{\nu}(\bar{\psi}\psi) 4.000 3.914 0 1 -1 \square(\bar{\psi}\psi) 4.000 4.105 2 1 1 T_{\mu\nu} 3.000 3.000 2 1 -1 \epsilon_{\rho\alpha\beta}\partial_{\alpha}T_{\mu\beta} 4.000 4.326 3 1 1 \partial_{\alpha}T_{\mu\nu} 4.000 3.763 Table S1: Scaling dimensions of the fields in the 3D gauged Majorana fermion CFT (we restrict to \Delta\leq 4 ). Fields are labeled by their angular momentum L , scaling dimension \Delta , and charge under the layer \mathbb{Z}_{2} symmetry and space-time parity. The latter is not directly accessible in our microscopic model on the fuzzy sphere. The data at N=14 shows good agreement with the exact scaling dimensions of the theory. SII S2. Gap analysis at the critical point In this section, we provide additional numerical evidence for the critical point studied in the main text. In particular, we perform finite-size scaling analysis of the gaps in the CFT and non-CFT sectors of the Hilbert space. SII.1 Non-CFT states in the neutral sector The complete spectrum of the model at the critical point (V_{0}^{\text{inter}}=0.48,h=0.58) in the neutral sector (i.e. 2Q=N-2 ) is shown in Fig. S1 . As detailed in the main text, we identify the ( N -even, \mathbb{Z}_{2} -even) and ( N -odd, \mathbb{Z}_{2} -odd) sectors as the CFT sectors, where we are able to accurately capture the operator content at low energies; the other layer-parity sectors are effectively decoupled from the phase transition. Figure S1: Finite-size spectrum of the model in the main text at its critical point, rescaled such that the stress-energy tensor has scaling dimension \Delta_{T}=3 . (a) The even-particle sector at system size N=14 . At low energy and angular momentum, the expected CFT scaling dimensions (shown as black lines) are in good agreement with the microscopic energies of the \mathbb{Z}_{2} -even sector. The \mathbb{Z}_{2} -odd sector is decoupled from the critical theory. The collective excitation branches of the bilayer FQH state are also highlighted by the grey box. (b) Same as (a) but for the odd-particle sector at N=13 . Similarly to the even-particle sector, the CFT tower of states is well captured by the \mathbb{Z}_{2} -odd sector, while the reverse parity sector does not participate to the transition. In this sector, the collective excitation corresponds to the unpaired fermion mode of the Moore–Read state Möller et al. 2011 ; Bonderson et al. 2011 . At higher angular momenta, the collective modes corresponding to gapped bilayer FQH states Renn and Roberts 1993 ; MacDonald and Zhang 1994 ; Shizuya 2003 can also be observed in Fig. S1 . In the N -even sector, we identify the “magnetoroton” branch Girvin et al. 1986 , corresponding to the density-wave distortion of the FQH droplet, while in the odd-particle sector we identify the Bogoliubov-type quasiparticle of the weakly paired states—the “unpaired fermion” branch Möller et al. 2011 ; Bonderson et al. 2011 . All excitations described above (which are not part of the CFT sector) are expected to display a finite gap in the thermodynamic limit, hence they will reside deep within the continuum of the spectrum whose low-energy part will be dominated by CFT states. In Fig. S2 , we verify this picture by performing thermodynamic-limit extrapolations for several of the expected gapped levels in the spectrum. This confirms that all collective modes (i.e., the magnetoroton and unpaired fermion in all layer \mathbb{Z}_{2} sectors) as well as the singlet states in the decoupled \mathbb{Z}_{2} sectors converge to a finite energy gap. Figure S2: Finite-size scaling of energy gaps in the neutral sector. (a) The even-particle sector for system sizes N=6-14 . The magnetoroton collective modes of the Halperin- 220 /Pfaffian states are identified as the lowest energy state outside the CFT tower region. The angular momentum values of these states are, as a function of system size, L_{\mathbb{Z}_{2}=+1}=(3,4,4,5,6) and L_{\mathbb{Z}_{2}=-1}=(2,3,3,3,4) , with gaps extrapolating to a finite value. In the \mathbb{Z}_{2} -odd sector, we also confirm that the lowest singlet state has a finite energy gap. (b) The odd-particle sector for system sizes N=7-13 . Here, the unpaired fermion collective modes are also identified, taking angular momentum values L_{\mathbb{Z}_{2}=-1}=(7/2,9/2,9/2,11/2) and L_{\mathbb{Z}_{2}=+1}=5/2 throughout (note that this does not hinder the identification of CFT state as it is in the decoupled sector). Similarly, the L=1/2 state also extrapolates to a finite gap. We note that the scenario in Fig. S2 is similar to Ref. Voinea et al. 2025 , where gapped collective excitations were constructed using a mean-field picture, allowing their identification in the spectrum even when they are no longer well-separated in energy from other eigenstates. This mean-field approach was facilitated by the fact that the charge sector remains in the same phase throughout the transition in Ref. Voinea et al. 2025 , while in the present case we have two distinct phases on either side of the transition. Nevertheless, one could attempt a similar identification here starting from the variational ansatz proposed in Ref. Ho 1995 . SII.2 Charge gap At the critical point, the U(1) electric charge is expected to decouple (albeit with a remnant topological contribution); we probe this by studying the behaviour of the charge gap as a function of the tunneling field. This is particularly important, as the charge excitations of the Halperin-220 and Pfaffian states are inherently different, with the latter carrying non-Abelian exchange statistics. Previously, a sharp upward cusp in the charge gap at the transition was experimentally observed in wide quantum wells at \nu=1/2 Suen et al. 1994 , and it has also been reproduced in numerical simulations on realistic models Nomura and Yoshioka 2004 ; Peterson and Das Sarma 2010 . Therefore, non-analytic behaviour is expected at the critical point, that would signal a transition between these distinct charged quasiparticles. Figure S3: (a) Evolution for the charge gap \Delta E_{c} as a function of tunneling, for a system of N=12 bosons. Two apparent transitions are encountered. The transition in the 2-quasiparticle state at flux 2Q-1 happens at h_{c}^{\text{(qp)}}\approx 0.465 , and the angular momentum sector changes from L=3 to L=2 , and one corresponding to the quasihole flux at 2Q+1 ; similarly, the quasihole transition happens at h_{c}^{\text{(qh)}}\approx 0.57 , closer to the critical field identified in the neutral sector, with the angular momentum changing from L=1 to L=4 . (b) Extrapolation of the charge gap for N=6-14 particles. While there are strong finite-size effects, the charge sector appears to have a non-zero gap in the thermodynamic limit. We consider the charge gap defined as \Delta E_{c}=f\big[\tilde{E}_{0}(N,2Q+1)+\tilde{E}_{0}(N,2Q-1)-2\tilde{E}_{0}(N,2Q)\big]/n_{q}\,. (S8) In this system, the charged excitations can only be created in pairs (each one contains half of a flux quantum), and therefore to calculate the energy of a single excitation we set n_{q}=2 . The energies are also rescaled by a factor f=\sqrt{2Q\nu/N} to account for the deviation from the filling factor \nu in the thermodynamic limit. Morf et al. 2002 . The corrected energies \tilde{E}_{0} are calculated from the ground state energies at a given flux E_{0} by including the correction from the neutralizing background charge, in terms of the quasiparticle number n_{q} , the quasiparticle charge e_{q} , and the average charging energy per particle pair C_{2Q} Jain and Kamilla 1997 ; Balram and Wójs 2020 : \tilde{E}_{0}(N,2Q)=E_{0}(N,2Q)-C_{2Q}(N^{2}-n_{q}^{2}e_{q}^{2})/2\,,\quad\mathrm{with}\quad C_{2Q}=V_{0}\frac{4Q+1}{(2Q+1)^{2}}\,. (S9) Here, the excitation charge is \pm e/4 , and we use the averaged pseudopotential V_{0}=(V_{0}^{\text{inter}}+V_{0}^{\text{intra}})/2 . Fig. S3 shows the behavior of the charged excitations across the phase transition. We observe two upward cusps, corresponding to transitions in the quasiparticle sector (i.e. \tilde{E}_{0}(N,2Q-1)-\tilde{E}_{0}(N,2Q) ) and the quasihole sector (i.e. \tilde{E}_{0}(N,2Q+1)-\tilde{E}_{0}(N,2Q) ). In finite-size systems, the quasihole, the quasiparticle and the neutral sector critical points do not coincide, which we attribute to finite-size effects; however, since the charged excitations of the Halperin 220 and Pfaffian states have specific countings in different angular momentum sectors Read and Rezayi 1996 , the ground state sectors at the first order transitions provide an additional signature of the Abelian to non-Abelian phase transition. For example, the 2-quasihole state of the Pffafian phase can only take angular momentum values L\in\{N/2,N/2-2,\dots\} in a particular system size N , which our model is in agreement with.