String theory builds a four-dimensional particle-physics model by curling up six extra dimensions into a Calabi–Yau manifold and placing an internal gauge field on it. Mathematically, that gauge field is a vector bundle. The resulting vacuum is supersymmetric and stable only if the bundle is slope-stable: no sub-piece of the bundle may carry an average "degree," measured against the chosen size parameters (the Kähler class), that is at least as large as the whole bundle's. The two smooth heterotic constructions in the literature that reproduce the Standard Model spectrum exactly both leave an extra gauged $U(1)_{B-L}$ symmetry at low energy. Here $B-L$ is baryon number minus lepton number. Low-energy model builders want this symmetry because, as long as it is unbroken, it enforces R-parity automatically. It must still break eventually, and the favored way is for a scalar neutrino (sneutrino) to acquire a vacuum expectation value (VEV).
In the string picture, giving a charged matter field a VEV deforms the bundle itself: the original bundle $V$ is glued to a trivial piece into a larger bundle $V'$ along a class in $H^1(X,V)$, the same cohomology group whose elements are the matter fields. The proposal is to build and run a verifier that takes a symmetry-breaking direction in the four-dimensional theory, constructs the corresponding deformed bundle, and decides whether $V'$ is stable, strictly semistable, or unstable across the Kähler cone. The physical question it answers is whether a given sneutrino-driven breaking of $B-L$ and R-parity can be a supersymmetric vacuum of the full compactification, or whether it can happen only once supersymmetry is broken. The second answer fixes the scale and size of the R-parity-violating couplings that result.
Core Claim
The simplest case already makes the claim sharp. A nonsplit extension $0\to V\to V'\to \mathcal{O}_X\to 0$, with $V$ stable and $c_1(V)=0$, contains $V$ as a sub-bundle of slope zero, which equals the slope of $V'$. So $V'$ is at best strictly semistable and has no Hermitian Yang–Mills connection. It is nevertheless typically simple, meaning $h^0(\mathrm{End}\,V')=1$. Simplicity is the diagnostic that the published heterotic standard models rely on, so in exactly the symmetry-breaking situation it would wrongly certify the deformed vacuum. This is the geometric shadow of an uncancelled four-dimensional D-term, the energy a charged field pays when nothing of opposite charge balances it.
The second part of the claim concerns the three-generation model of A heterotic standard model. That model has right-handed neutrinos and, according to its authors, no exotic matter, so the right-handed sneutrino has no partner of opposite $B-L$ charge to cancel its D-term. We conjecture, and propose to test, that every supersymmetric deformation which breaks $B-L$ there must involve several fields jointly. Pure right-handed-sneutrino breaking would then be possible only at the supersymmetry-breaking scale.
If that holds, the effective bilinear coupling $\kappa_i L_i H_u$ is not a generic free parameter. It is induced as $\kappa_i \sim y_\nu\langle\tilde\nu^c\rangle$, where $y_\nu$ is the neutrino Yukawa coupling and $\langle\tilde\nu^c\rangle$ is tied to soft masses. This sits in direct tension with R-parity conservation from a top down perspective, which finds bilinear R-parity violation generic and fine-tuned in heterotic orbifolds and $G_2$ compactifications but explicitly does not treat smooth Calabi–Yau manifolds with non-abelian bundles. What is new is a computable criterion linking four-dimensional breaking directions to bundle stability, with a Kähler-class dependence that tree-level four-dimensional D-flatness for a non-anomalous $U(1)$ cannot see.
Research Program
The work sits on two existing constructions, both of which are motivated by and constrained by prior papers. The first is the rank-4 extension bundle on the torus-fibered threefold with $\mathbb{Z}_3\times\mathbb{Z}_3$ fundamental group, built in Vector Bundle Extensions, Sheaf Cohomology, and the Heterotic Standard Model. Its moduli census in Heterotic standard model moduli counts 19 observable bundle moduli and 3 Kähler moduli, and verifies only $H^0(X,V\otimes V^\vee)=1$. The second is the non-positive monad on the bi-cubic, from Exploring positive monad bundles and a new heterotic standard model. There stability of $V$ is proven, but the realistic spectrum appears only at a special locus. Both papers leave the dynamics of $B-L$ breaking unmodeled. We take the bundles as given and study their deformations.
The first step is a dictionary between four-dimensional fields and extension classes. Each field, such as the right-handed sneutrino $\tilde\nu^c_i$, the slepton doublet $\tilde L_i$, or the Higgs fields, corresponds to a class in a specific $H^1$ group of the covering space with a specific character under the discrete group. For the BHOP bundle, the Leray spectral-sequence machinery of the extension paper supplies these classes. For the monad, the Koszul-sequence machinery of Heterotic compactification, an algorithmic approach supplies them. A VEV along a single field produces an equivariant extension. A VEV along several fields with different charges produces an iterated extension or a deformation of a split bundle, which mixes both gluing directions.
The second step is the stability decision itself. We generalize the Hoppe-type reduction used by Anderson, He and Lukas to Picard number greater than one. The Kähler cones here have dimension 3 and 2. For each line bundle $\mathcal{L}$ in a finite window, we compute $h^0(X,\wedge^p V'\otimes\mathcal{L})$ for $0<p<\mathrm{rk}\,V'$. The window is bounded because a destabilizing sub-sheaf must have first Chern class lying between the stability walls of the cone. We record which classes $c_1(\mathcal{L})$ can destabilize, and then compute the sub-region of the Kähler cone in which no such sub-sheaf has slope at least that of $V'$. The output for each breaking direction is a stability region: empty, the whole cone, or a chamber bounded by walls.
The third step feeds that map back into phenomenology. Directions that are stable somewhere define supersymmetric $B-L$ breaking. Their R-parity-violating couplings follow from the mixing formulas in Higgs sector and R-parity breaking couplings in models with broken U(1)B−L gauge symmetry, which derives lepton–gaugino–Higgsino mixing from a sneutrino VEV. Directions that are only semistable force breaking to the soft scale. In both cases the induced $\kappa_i$ and left-sneutrino VEVs are confronted with the bound $v_\tau<6$ GeV at 1 TeV gaugino mass reported in that paper.
Baselines
The naive reference point is the simplicity check alone: compute $h^0(\mathrm{End}\,V')$ for each deformed bundle and declare the deformation acceptable when it equals one. This is exactly what the foundational heterotic standard model papers did for the undeformed bundle, so it is the method the community would reach for first. We run it on the same deformed bundles, with the same Macaulay2 and spectral-sequence code and the same machine budget. The proposal beats it if we exhibit deformations that simplicity accepts but our verifier proves strictly semistable or unstable. The single-field extension argument above predicts that such cases exist. Producing them explicitly, with the destabilizing sub-sheaf named, is the decisive output.
The strongest competitor is a purely four-dimensional D-flatness analysis. It uses the known $SU(3)\times SU(2)\times U(1)_Y\times U(1)_{B-L}$ charges of the zero modes and asks which directions are D-flat, via holomorphic gauge-invariant monomials. It also checks F-flatness against the known cubic couplings. This is the strongest baseline because, for small VEVs and a non-anomalous $U(1)$, it should agree with geometry. It is cheap, and it is what low-energy model builders already do. We run it on exactly the same field content and the same list of breaking directions.
The proposal "beats" it only in one of two ways. Either the verdict depends on the Kähler class, which tree-level four-dimensional D-terms cannot reproduce, or the 19 uncharged bundle moduli change the verdict in a way the charge analysis misses. If the two always agree, we will report that the geometric verifier is a rigorous all-orders certificate of the four-dimensional answer rather than a source of new constraints. No reported result found on Lacuna for either baseline applied to deformed, symmetry-breaking bundles.
As ground truth for the verifier itself, we use the classification in Heterotic compactification, an algorithmic approach. It reports that all 37 classified positive monad bundles are proven stable. Our generalized pipeline must reproduce stability on all 37 undeformed bundles before it is trusted on anything else.
Two ablations isolate the key ingredients. The first drops equivariance and checks stability on the covering space only. The second evaluates stability at a single interior Kähler point rather than mapping the whole cone. If either ablation leaves the verdicts unchanged across both models, that ingredient is not doing work.
Why not just…?
Why not just trust $h^0(\mathrm{End}\,V)=1$, as the original papers did? Braun, He, Ovrut and Pantev themselves present simplicity as necessary but not sufficient. Extremal Bundles on Calabi–Yau Threefolds shows that bundles with the right invariants still pose a separate stability problem. Triadophilia: A Special Corner of the Landscape shows that stability does not persist automatically across geometric transitions. For the deformed bundles relevant to symmetry breaking the situation is worse: a nonsplit extension by the trivial bundle can be simple and still have no Hermitian Yang–Mills connection. The first experiment below is designed to make that failure explicit.
Why not just do the D-flatness analysis in four dimensions and skip the geometry? This is the real threat, and we do not argue around it. For a non-anomalous $U(1)_{B-L}$ at small VEV, the two analyses should coincide, and in that regime the geometric work is a consistency check. The geometry adds information in three places: for VEVs comparable to the compactification scale, where the effective theory stops being reliable; through dependence on the Kähler class, which tree-level four-dimensional D-terms lack; and through mixing with uncharged bundle moduli. The comparison experiment quantifies how often each of these occurs. If the answer is "never," that is the reported outcome.
Why not just impose R-parity by hand, or study gauged $B-L$ purely bottom-up? Discrete global symmetries are expected to be violated by quantum gravity. The bottom-up gauged models of Puolamäki and of “L = R” — U(1) R as the origin of leptonic ‘RPV’ leave the breaking scale and the bilinear couplings as inputs. The top-down orbifold analysis reaches the opposite conclusion, that bilinear R-parity violation is generic and fine-tuned. Only an ultraviolet construction with an explicit spectrum, which is what the two smooth models provide, can decide which picture applies to smooth Calabi–Yau vacua.
Why not just train a neural network to find the Hermitian Yang–Mills connection and read stability off its failure? Machine learning line bundle connections reaches within 5% on the quintic, but only for abelian line bundles. A semistable but non-polystable bundle would show up as slow, ambiguous non-convergence rather than a clean verdict. Classifiers trained on CICY data, such as Machine learning CICY threefolds, reach F-values of only about 0.26 to 0.4 even on discrete-symmetry labels. We treat numerical connections as an optional cross-check on the one or two deformations the algebraic verifier certifies, not as a replacement for it.
Proposed Experiments
Verifier validation. The verifier must reproduce stability for all 37 positive monads of Anderson, He and Lukas, and for the bi-cubic bundle. It must also correctly label two synthetic families. The first is single-field extensions of these bundles by $\mathcal{O}_X$, which are provably strictly semistable. The second is two-sided deformations of split polystable sums $V_1\oplus V_2$, where both gluing directions are switched on together and the result should be stable in the expected chamber. The metric is exact agreement with these known labels, together with wall-clock time as a function of the Kähler-cone dimension and the line-bundle window.
The decisive check. For the BHOP bundle and the bi-cubic monad, we enumerate breaking directions built from single and multiple VEVs among $\tilde\nu^c_i$, $\tilde L_i$, $H_u$ and $H_d$. For each direction we record the simplicity verdict, the four-dimensional D- and F-flatness verdict, and the geometric stability region in the Kähler cone. The headline quantities are three. The first is the number of directions where simplicity passes but stability fails. The second is the number where the four-dimensional and geometric verdicts disagree. The third is the fraction of geometrically allowed directions whose allowed region is a proper chamber rather than the full cone.
Hypothesis test. The conjecture is that pure right-handed-sneutrino breaking is never supersymmetric in the BHOP model. It is refuted if any single-$\tilde\nu^c$ deformation has a non-empty stability region.
Phenomenological follow-through. For the surviving directions, we compute the induced $\kappa_i$ and left-sneutrino VEVs, and the R-parity-violating trilinear couplings, using Puolamäki's mixing relations. We then check them against $v_\tau<6$ GeV. The result settles, for these two smooth vacua, whether the top-down expectation of generic, fine-tuned bilinear R-parity violation carries over.
References
- A heterotic standard model
- R-parity conservation from a top down perspective
- Vector bundle extensions, sheaf cohomology, and the heterotic standard model
- Heterotic standard model moduli
- Exploring positive monad bundles and a new heterotic standard model
- Heterotic compactification, an algorithmic approach
- Higgs sector andR-parity breaking couplings in models with broken<mml:math xmlns:mml="http://www.w3.org/1998/Math/MathML" display="inline"><mml:mrow><mml:msub><mml:mrow><mml:mi>U</mml:mi><mml:mo>(</mml:mo><mml:mn>1</mml:mn><mml:mo>)</mml:mo><mml:mn/></mml:mrow><mml:mrow><mml:mi>B</mml:mi><mml:mi>−</mml:mi><mml:mi>L</mml:mi></mml:mrow></mml:msub></mml:mrow></mml:math>gauge symmetry
- Extremal Bundles on Calabi–Yau Threefolds
- Triadophilia: A Special Corner of the Landscape
- “L = R” — U(1) R as the origin of leptonic ‘RPV’
- Machine learning line bundle connections
- Machine learning CICY threefolds