index

Binding Energy, Critical Radii, & Information Maintenance Tax

· 15min

Every atom in your body pays continuous energy to maintain its structure against entropy. Nuclear binding energy, chemical bonds, gravitational self-energy—these represent ongoing thermodynamic costs, an information maintenance tax that all organized systems pay to resist dissolution into thermal chaos.

White dwarfs accreting toward the Chandrasekhar limit reveal the mechanism. Geometric gravitational compression increases by a factor of 2.2 while organizational complexity explodes by a factor of 2200—a 1000-fold disparity suggesting information bankruptcy drives collapse. Independent analysis of two white dwarf catalogs identifies R/RS103R/R_S \approx 10^3 as a discrete phase transition boundary: objects crossing this threshold appear systematically older than mass-and-temperature matched references, with the Montreal sample showing +103 Myr excess aging at 3.59σ significance and Gaia confirming at 34.6σ. The effect concentrates below R/RS1000R/R_S \approx 1000 and vanishes above, consistent with sector saturation forcing accelerated entropy production. The correspondence between binding energies and thermodynamic maintenance costs spans 57 orders of magnitude, from elementary particles to black holes.

The Thermodynamic Tax

Landauer’s principle establishes the minimum energy cost for information maintenance against thermal noise—kBTln2k_B T \ln 2 per bit, where kBk_B is Boltzmann’s constant and TT is temperature1. Binding energy across all scales represents this ongoing thermodynamic maintenance against entropy. Rather than stored or latent energy that could be released, binding energy quantifies the continuous work required to keep structure coherent. Nuclear binding maintains quark confinement, electromagnetic binding maintains atomic configurations, gravitational binding maintains bulk coherence—each through sustained energy expenditure against thermal randomization.

Any organized system requires bits to specify its configuration. For a bound system with NN particles of mass mm confined to radius RR with characteristic momentum pp, the bit count equals,

Nbits=log2(Ωh3N),N_{\text{bits}} = \log_2\left(\frac{\Omega}{h^{3N}}\right),

where Ω\Omega is the accessible phase space volume and hh is Planck’s constant. Each particle contributes (Rp/)3\sim (Rp/\hbar)^3 states, so the total scales as,

Ω(Rp)3N,\Omega \sim \left(\frac{Rp}{\hbar}\right)^{3N},

where pGMm/Rp \sim \sqrt{GMm/R} for gravitationally bound systems at virial equilibrium.

Thermal and quantum fluctuations constantly randomize configurations, requiring continuous energy expenditure,

Em=Nbits×kBTln2.E_m = N_{\text{bits}} \times k_B T \ln 2.

This maintenance energy equals binding energy for the dominant force at each scale. For gravitational systems, equating EmE_m with binding energy GM2/RGM^2/R and incorporating the Landauer-Bekenstein-Hawking factor of 2 from dimensional reduction at horizons yields,

EmMc2=RSR,\frac{E_m}{Mc^2} = \frac{R_S}{R},

where RS=2GM/c2R_S = 2GM/c^2 is the Schwarzschild radius. This creates a complete energy budget partition,

Em+Ea=Mc2,E_m + E_a = Mc^2,

where Ea=Mc2(1RS/R)E_a = Mc^2(1 - R_S/R) represents energy available for work beyond maintenance.

The Geometric Origin of Maintenance

The thermodynamic tax has a geometric origin. Any organized system enforces feasibility constraints on its dynamics: electron shells confine electrons to quantized orbitals, nuclear binding confines quarks within hadrons, gravitational virial relations confine matter within characteristic radii. These constraints project the system’s dynamics onto admissible configurations.

When constraints are state-independent—the same restriction everywhere in phase space—the projected dynamics can still derive from a scalar potential. The system descends toward equilibrium along a well-defined gradient. When constraints vary with state—when the admissible directions depend on where the system currently sits—the projection generically introduces curl into the effective dynamics.

Curl measures irreducible circulation: work that must be continuously supplied because no global potential exists. The curl-maintenance functional,

Mcurl=12dα2dV,\mathcal{M}_{\mathrm{curl}} = \frac{1}{2} \int |d\alpha|^2 \, dV,

where α=F\alpha = F^\flat is the 1-form dual to the correction field FF, quantifies this cost. On compact manifolds with trivial first cohomology, a spectral lower bound ensures that non-integrable projections carry an irreducible maintenance floor proportional to the projection defect magnitude.

Binding energies are curl-maintenance costs—the thermodynamic price of enforcing non-integrable constraints against entropy. Nuclear binding maintains quark confinement against QCD fluctuations. Electromagnetic binding maintains electron configurations against thermal noise. Gravitational binding maintains bulk coherence against dispersal. Each represents continuous energy expenditure to sustain state-dependent feasibility projections that lack global potentials.

Force-Specific Bankruptcy Radii

Each fundamental force imposes a characteristic bankruptcy—the scale where maintenance costs equal total energy. For the strong force at the QCD confinement scale,

rQCD=ΛQCD1 fm,r_{\text{QCD}} = \frac{\hbar}{\Lambda_{\text{QCD}}} \approx 1 \text{ fm},

where ΛQCD200\Lambda_{\text{QCD}} \approx 200 MeV sets quark confinement energy. Electromagnetic binding reaches bankruptcy at the classical electron radius,

rEM=ke2mec22.8×1015 m,r_{\text{EM}} = \frac{ke^2}{m_ec^2} \approx 2.8 \times 10^{-15} \text{ m},

where kk is Coulomb’s constant, ee is elementary charge, mem_e is electron mass, and cc is light speed. Gravitational systems fail at,

rgrav=RS=2GMc2.r_{\text{grav}} = R_S = \frac{2GM}{c^2}.

The maintenance fraction ξ\xi quantifies the fraction of a system’s energy budget devoted to curvature maintenance (§5 of Triadic Tension, Decade Symmetry, & Dissipation Flow). Observed values cluster near decade-spaced plateaus: elementary particles at ξ106\xi \sim 10^{-6}, atoms at 103\sim 10^{-3}, molecules at 102\sim 10^{-2}, and biological systems at 101\sim 10^{-1}. These values are order-of-magnitude estimates from the characteristic energy scales at each organizational level—the fine structure constant α25×105\alpha^2 \approx 5 \times 10^{-5} sets the electromagnetic coupling scale, Bohr-to-nuclear radius ratios set atomic coordination costs, and conformational entropy sets molecular overhead. The decade spacing itself is a prediction of the RG flow with C10C_{10} symmetry (§4 of the same post), where each RG period spans one decade in scale.

Forcercritr_{\text{crit}}Typical r/rcritr/r_{\text{crit}}Maintenance Cost
QCD1 fm~100~1% rest mass
EM2.8×10152.8 \times 10^{-15} m~10510^5~10510^{-5} rest mass
Grav2GM/c22GM/c^2variesRS/R×R_S/R \times rest mass

QCD and electromagnetic forces operate far from bankruptcy, enabling stable structures with minimal overhead. Gravitational systems span the entire spectrum from negligible costs in atoms to complete dedication at black hole horizons. These bankruptcy radii mark where accumulated curvature has consumed all available capacity—the system has exhausted integrable degrees of freedom, and maintenance cost equals total energy budget. At this threshold, no further organizational complexity can be sustained without catastrophic reorganization.

Information Bankruptcy Trajectory

White dwarfs accreting mass toward the Chandrasekhar limit MCh=1.36MM_{\text{Ch}} = 1.36 M_{\odot}2 provide a natural laboratory for studying information bankruptcy. The complete trajectory reveals that organizational complexity determines collapse.

The complexity multiplier quantifying the overhead beyond baseline requirements follows,

M(ξ,D)=φ2D2×(1ξ)u,M(\xi,\mathcal{D}) = \varphi^{2^{\mathcal{D}-2}} \times (1-\xi)^{-u^*},

where φ=(1+5)/2\varphi = (1+\sqrt{5})/2 is the golden ratio and u=4πφ2/103.29u^* = 4\pi\varphi^2/10 \approx 3.29 is the coupling constant derived in the constraint geometry framework. This contains two competing terms. The dimensional factor φ2D2\varphi^{2^{\mathcal{D}-2}} decreases mildly as effective dimension D\mathcal{D} drops from 3 toward 2, representing reduced interference in lower dimensions. The bankruptcy factor (1ξ)u(1-\xi)^{-u^*} diverges catastrophically as maintenance fraction ξ\xi approaches unity.

Numerical integration from stable white dwarfs through collapse yields the trajectory (using constant radius R5000R \approx 5000 km from electron degeneracy pressure):

MM (M)(M_{\odot})RS/RR_S/Rξ\xiD\mathcal{D}φ2D2\varphi^{2^{\mathcal{D}-2}}(1ξ)u(1-\xi)^{-u^*}M(ξ,D)M(\xi,\mathcal{D})ξ×M\xi \times MStatus
0.603.6×1043.6 \times 10^{-4}0.0662.972.611.243.20.21Stable
1.006.0×1046.0 \times 10^{-4}0.272.872.522.907.32.0Normal
1.177.0×1047.0 \times 10^{-4}0.462.782.425.6613.76.3Anomaly
1.307.8×1047.8 \times 10^{-4}0.632.702.3512.429.118.3Critical
1.358.0×1048.0 \times 10^{-4}0.972.532.15229492477Collapse

The trajectory reveals the mechanism. Geometric compression RS/RR_S/R increases by a factor of 2.2 from M=0.60M = 0.60 to 1.35M1.35 M_{\odot}—mild gravitational strengthening. Meanwhile, organizational complexity ξ×M\xi \times M explodes by a factor of 2200 over the same range. This 1000-fold disparity indicates that information bankruptcy, not gravitational compression alone, drives instability. The dimensional factor φ2D2\varphi^{2^{\mathcal{D}-2}} drops modestly from 2.61 to 2.15 as D\mathcal{D} flows from 2.97 to 2.53—barely 20% variation. The bankruptcy factor (1ξ)u(1-\xi)^{-u^*} generates the explosion: from 1.24 at stable masses to 229 near collapse, a 185-fold increase driven by the same coupling constant u3.29u^* \approx 3.29 that governs the β\beta-function’s RG flow.

Basin of Attraction Entrance

The observational anomaly at R/RS=103R/R_S = 10^3 corresponds to M1.17MM \approx 1.17 M_{\odot} where ξ=0.46\xi = 0.46 and (1ξ)u=5.66(1-\xi)^{-u^*} = 5.66. This marks the boundary where thermodynamic bankruptcy becomes inevitable rather than merely possible—the entrance to the basin of attraction toward organizational collapse. The basin language is precise: as formalized in the feasibility projection framework, it suffices that the system’s dynamics land within the basin of attraction of the feasibility operator Πx\Pi_x, with exact feasibility enforced by the projector rather than the proposal dynamics3. Below R/RS103R/R_S \approx 10^3, proposals that would maintain the current organizational state no longer converge under the feasibility map—the system has exited the projector’s attraction basin, and no correction schedule can restore convergence without structural reorganization. The residual-gradient correction framework from constrained diffusion for ACOPF3 provides the operational analog: once the infeasibility residual R(x)\mathcal{R}(x) exceeds the basin boundary, gradient corrections x=xλxR(x)x' = x - \lambda\,\nabla_x \mathcal{R}(x) no longer converge to a feasible state, and the system must jump to a qualitatively different organizational mode.

Before this threshold, complexity overhead grows slowly—a factor of 3.6 from M=0.6M = 0.6 to 1.17M1.17 M_{\odot}. After crossing R/RS=103R/R_S = 10^3, overhead explodes—a factor of 36 from M=1.17M = 1.17 to 1.35M1.35 M_{\odot}. The (1ξ)u(1-\xi)^{-u^*} value of 5.66 at the anomaly threshold represents the onset of nonlinear divergence. Systems maintain ξ<0.5\xi < 0.5 through moderate overhead factors. Beyond ξ0.5\xi \sim 0.5, divergence accelerates uncontrollably.

Analysis of 7,515 Gaia DR3 white dwarfs4 reveals that objects in the anomaly zone (R/RS[700,1000)R/R_S \in [700, 1000)) appear systematically older than mass- and temperature-matched references, with a mean excess of +76+76 Myr (14.5σ14.5\sigma, Wilcoxon p=4.27×1084p = 4.27 \times 10^{-84}, 95% bootstrap CI: [+67,+85][+67, +85] Myr). Binned analysis reveals a monotonic compactness gradient: +345+345 Myr at R/RS=500R/R_S = 500700700 (using relativistic ONe-core tracks for the ultra-massive subpopulation), +108+108 Myr at 700700900900, +6+6 Myr at 90090011001100, and negligible anomaly above R/RS=1100R/R_S = 1100. The full analysis, including dual methodology for different mass regimes and systematic checks on core composition, is developed in Repair as Local Optimization in Constraint Geometry.

The discrete, threshold-like character of the effect—rather than a continuous gradient with mass—suggests a phase transition in constraint geometry. Once a white dwarf crosses into the R/R_S < 1000 regime, it enters a qualitatively different organizational basin characterized by accelerated entropy production and faster evolution through state space. This is consistent with sector saturation in the constraint eigenvalue framework: at R/R_S ≈ 10³, one organizational sector reaches capacity, forcing reconfiguration whose thermodynamic cost manifests as ~100 Myr of accelerated aging.

Independent observational work has recently identified anomalous behavior in ultra-massive white dwarfs concentrated near the same compactness scale. A 2026 analysis of Gaia-selected white dwarfs along the Q branch5 reports extended cooling delays localized to this regime, inferred from kinematic clustering and photometric evolution rather than direct age residuals. While interpreted within conventional stellar-evolution frameworks, the discreteness and localization of the effect closely mirror the threshold behavior identified here. The convergence of independent observables—cooling dwell time on the Q branch and apparent age excess relative to matched controls—suggests a structural transition in white dwarf evolution rather than a smooth modification of cooling physics, reinforcing the interpretation of R/RS103R/R_S \sim 10^3 as a physically meaningful regime boundary across independent observational parameterizations rather than a model-dependent artifact.

The Discontinuous Jump to Neutron Degeneracy

White dwarfs do not smoothly flow to the (ξ=1,D=2)(\xi=1, \mathcal{D}=2) black hole fixed point. Instead, information bankruptcy forces a discontinuous organizational jump. At MMChM \approx M_{\text{Ch}}, the system reaches ξ0.97\xi \approx 0.97, D2.5\mathcal{D} \approx 2.5 with maintenance cost ξ×M477\xi \times M \approx 477. This overhead exceeds sustainable levels, triggering catastrophic reorganization—the white dwarf jumps to neutron degeneracy at ξ0.3\xi \sim 0.3, D2.5\mathcal{D} \sim 2.5 with complexity ξ×M2.3\xi \times M \approx 2.3.

The organizational complexity drops by a factor of 207, requiring massive information restructuring. The energy cost of this reorganization follows from Landauer’s principle: counting the bits required to reorganize phase space information from electron to neutron degeneracy (ΔNbits4.5×1058\Delta N_{\text{bits}} \approx 4.5 \times 10^{58}) at the shock temperature T109T \sim 10^9 K gives a transition energy of Etrans=4.3×1044E_{\text{trans}} = 4.3 \times 10^{44} J—matching observed Type Ia supernova energies6 to within measurement uncertainty. The full derivation, which requires only four observational inputs (Chandrasekhar mass, white dwarf radius, neutron star radius, and shock temperature) and no parameters from the constraint geometry, is developed in the companion paper. The white dwarf collapses because maintaining organizational complexity at ξ0.97\xi \approx 0.97 requires infinite energy through the (1ξ)u(1-\xi)^{-u^*} divergence.

The (ξ=1,D=2)(\xi=1, \mathcal{D}=2) black hole state requires additional compression beyond neutron star density, achievable only by exceeding the Oppenheimer-Volkoff limit7. White dwarf collapse represents information bankruptcy at electron degeneracy—a different failure mode with lower organizational overhead than neutron star collapse to black holes. This mechanism operates across all scales: ionization at atomic bankruptcy (~100 eV per particle), molecular dissociation at chemical limits (10 eV per bond), stellar collapse at gravitational thresholds (104410^{44} J for solar masses).

Information Processing Hierarchy

Black holes occupy the IR fixed point of the β\beta-function at (ξ=1,D=2)(\xi = 1, \mathcal{D} = 2), where all available energy is dedicated to maintaining horizon structure—maximum entropy as pure curvature maintenance with zero excess capacity8. The dimensional reduction to D=2\mathcal{D} = 2 at this fixed point is consistent with the holographic scaling of Bekenstein-Hawking entropy with boundary area rather than volume. All other organized systems operate at ξ<1\xi < 1, with residual capacity (1ξ)(1-\xi) available for organizational dynamics beyond pure pattern preservation.

Recent observations of the tidal disruption event AT2020afhd provide an empirical window into this regime9. The system exhibits synchronized X-ray and radio oscillations persisting over many dynamical times, consistent with Lense–Thirring precession driven by black hole spin. Angular momentum information remains coherently encoded in boundary geometry across extended timescales—behavior consistent with the information maintenance saturation expected as ξ1\xi \to 1. That such coherence emerges naturally in the strong-field limit suggests the dimensional reduction and boundary processing architecture may be generic features of gravitational information storage.

Critical mass phenomena mark bankruptcy thresholds. White dwarfs approaching the Chandrasekhar limit see the maintenance fraction diverge through the saturating form,

ξ(M)=ξ0(1M/MCh)u1+ξ0(1M/MCh)u,\xi(M) = \frac{\xi_0(1 - M/M_{\text{Ch}})^{-u^*}}{1 + \xi_0(1 - M/M_{\text{Ch}})^{-u^*}},

with ξ0=0.01\xi_0 = 0.01 and u=3.29u^* = 3.29. This prevents unphysical ξ>1\xi > 1 while capturing divergence as MMChM \to M_{\text{Ch}}. Combined with the complexity multiplier, this generates the observed 2200-fold increase in organizational overhead over the mass range from stable white dwarfs to collapse. Conversely, black holes face inverted criticality through Hawking evaporation8,

TH=c38πGMkB1M,T_H = \frac{\hbar c^3}{8\pi G M k_B} \propto \frac{1}{M},

where THT_H is Hawking temperature. Below approximately 101210^{12} kg, rising temperature creates thermal noise overwhelming error correction, triggering runaway evaporation.

The Deeper Correspondence

Every bound system—from nuclei to galaxies—pays continuous energy to maintain informational coherence, with cost scaling as Em=Mc2×(rcrit/r)×ξE_m = Mc^2 \times (r_{\text{crit}}/r) \times \xi. Each fundamental force acts as a different collection mechanism for the same underlying tax, with characteristic bankruptcy radii where maintenance equals available energy.

The geometric foundation connects this observation to the constraint geometry framework. Curl explains why maintenance is required: state-dependent feasibility projections introduce irreducible circulation that demands continuous energy supply. The coupling constant u=4πφ2/103.29u^* = 4\pi\varphi^2/10 \approx 3.29 explains when maintenance overwhelms the system: as ξ\xi approaches the bankruptcy regime, the complexity multiplier (1ξ)u(1-\xi)^{-u^*} diverges, integrable degrees of freedom are progressively exhausted, and the system must reorganize or collapse. Bankruptcy radii mark where accumulated curvature has consumed all available capacity.

The white dwarf trajectory through the basin of attraction threshold at R/RS=103R/R_S = 10^3 confirms the predicted mechanism, and the Type Ia supernova energy of 4.3×10444.3 \times 10^{44} J from Landauer bit-counting matches observation. Black holes represent the thermodynamic endpoint where all energy maintains horizon structure—maximum entropy as pure maintenance with zero excess capacity. Critical mass phenomena from Chandrasekhar to Oppenheimer-Volkoff limits7 emerge as bankruptcy thresholds where organizational modes become unsustainable.

Limitations and Falsifiability

The binding-as-maintenance interpretation rests on two empirical claims and one theoretical identification. If the white dwarf cooling anomaly at R/RS103R/R_S \approx 10^3 is explained by conventional physics—enhanced neutrino cooling, modified equations of state near the Chandrasekhar limit, or systematic biases in progenitor populations—then the sector saturation interpretation loses its primary evidence, though the Landauer energy calculation would remain independent. If the Type Ia supernova energy match (4.3×10444.3 \times 10^{44} J) is coincidental rather than causal—if the agreement between Landauer bit-counting and observed energies reflects fortuitous cancellation of errors in the four input parameters—then the framework’s strongest quantitative prediction dissolves. The theoretical identification of binding energy with curl-maintenance cost depends on the constraint geometry’s claim that state-dependent projections introduce irreducible circulation (Theorem 4 of the self-correction paper); if that theorem’s hypotheses fail on the relevant physical manifolds, the geometric foundation weakens. The methodology appendix below details the empirical analysis and its limitations.

Appendix: White Dwarf Anomaly Methodology

The empirical analysis supporting the R/RS103R/R_S \approx 10^3 threshold uses 7,515 white dwarfs from the Gaia DR3 catalog4, pre-filtered for R/RS[500,1500]R/R_S \in [500, 1500] and M1.1MM \geq 1.1 \, M_\odot (hydrogen-atmosphere, thick-envelope fits). Cooling ages are derived from Montréal DA thick-hydrogen-envelope tracks (0.200.201.30M1.30 \, M_\odot), supplemented by La Plata relativistic ONe-core and CO-core GR tracks extending coverage to 1.369M1.369 \, M_\odot and 1.382M1.382 \, M_\odot respectively.

Two statistical methods address different mass regimes. For R/RS[700,1000)R/R_S \in [700, 1000) versus the reference zone [1000,1500][1000, 1500], kk-nearest-neighbor matching (k=10k = 10) in standardized (M,logTeff)(M, \log T_{\mathrm{eff}}) space constructs cooling-age residuals Δti=tit~ref\Delta t_i = t_i - \tilde{t}_{\mathrm{ref}}, where Δt>0\Delta t > 0 indicates the anomaly-zone object appears older. For R/RS[500,700)R/R_S \in [500, 700), where all objects have M>1.32MM > 1.32 \, M_\odot and no reference population exists at comparable mass, a direct Mann-Whitney comparison against the 700–900 bin replaces kNN matching.

The anomaly shows a monotonic compactness gradient: +345+345 Myr median excess at 500–700 (p=1.5×106p = 1.5 \times 10^{-6}), +108+108 Myr at 700–900 (p=4.6×1029p = 4.6 \times 10^{-29}), +6+6 Myr at 900–1100, and null above 1300. Globally across 700–1000, the mean residual is +76+76 Myr (14.5σ14.5\sigma, Wilcoxon p=4.27×1084p = 4.27 \times 10^{-84}). The Spearman correlation ρ(M2,Δt)=+0.350\rho(M^2, \Delta t) = +0.350 (p=4.93×1030p = 4.93 \times 10^{-30}) confirms that more compact objects show larger positive residuals. The result is robust to core-composition assumptions: CO-core tracks yield +697+697 Myr for the 500–700 bin (p=5.9×1010p = 5.9 \times 10^{-10}). Full methodological details, including track construction, interpolation procedures, and systematic checks, are given in Repair as Local Optimization in Constraint Geometry.

High-mass white dwarfs may have systematic differences in progenitor populations or formation channels, and the causal mechanism cannot yet be distinguished from alternative explanations such as enhanced neutrino cooling, crystallization energy release, or GR corrections to the cooling rate near the Chandrasekhar limit.

Footnotes

  1. Landauer, R. (1961). Irreversibility and heat generation in the computing process. IBM Journal of Research and Development, 5(3), 183-191.

  2. Chandrasekhar, S. (1931). The maximum mass of ideal white dwarfs. Astrophysical Journal, 74, 81-82.

  3. Shekhar, S., Karn, A., Keshav, K., Bansal, S., & Pareek, P. (2026). “Fast Diffusion with Physics-Correction for ACOPF.” arXiv

    .03020. https://arxiv.org/abs/2602.03020 2

  4. Dufour, P., et al. (2017). The Montreal White Dwarf Database. Proceedings of the 20th European White Dwarf Workshop, 509, 3. See also Gentile Fusillo, N. P., et al. (2021). A catalogue of white dwarfs in Gaia EDR3. Monthly Notices of the Royal Astronomical Society, 508(3), 3877-3896. 2

  5. Ould Rouis, L. B., Hermes, J. J., Guidry, J. A., et al. (2026). White Dwarf Merger Remnants with Cooling Delays on the Q Branch Lack Strong Magnetism. Astrophysical Journal, submitted. arXiv

    .02670.

  6. Hillebrandt, W., & Niemeyer, J. C. (2000). Type Ia supernova explosion models. Annual Review of Astronomy and Astrophysics, 38(1), 191-230.

  7. Oppenheimer, J. R., & Volkoff, G. M. (1939). On massive neutron cores. Physical Review, 55(4), 374. 2

  8. Hawking, S. W. (1975). Particle creation by black holes. Communications in Mathematical Physics, 43(3), 199-220. 2

  9. Pasham, D. R., et al. (2025). Synchronized X-ray and radio variability from the tidal disruption event AT2020afhd consistent with Lense–Thirring precession. Science Advances, 11(49), eady9068.