pith. sign in

arxiv: 2607.00078 · v1 · pith:PVU35HBKnew · submitted 2026-06-30 · ✦ hep-th

Exploring Line Bundle Standard Models with Transformers

Jacky H. T. Yip , Alessandro Mininno , Gary Shiu This is my paper

Pith reviewed 2026-07-02 18:31 UTC · model grok-4.3

classification ✦ hep-th
keywords heterotic stringsline bundlesCalabi-Yau threefoldsreinforcement learningtransformersstandard modelanomaly cancellationpoly-stability
0
0 comments X

The pith

A Transformer-based reinforcement learning agent identifies line bundle sums on Calabi-Yau threefolds that satisfy the constraints for heterotic SU(5) standard models.

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The paper introduces LB-Explorer, a Transformer reinforcement learning architecture, to search for heterotic line bundle standard models arising from compactifications on smooth Calabi-Yau threefolds. It targets E8 x E8 vacua with SU(5) symmetry that can be broken to the Standard Model gauge group by discrete Wilson lines. The system is tested on complete intersection Calabi-Yau manifolds and learns to enforce the full set of physical requirements on the line bundle sums. Candidates are then passed to hybrid solvers for remaining exact conditions. The approach is presented as a scalable method for exploring large moduli spaces in the string landscape.

Core claim

The LB-Explorer efficiently learns constraints on the line bundle sums, guaranteeing the E8 gauge embedding, anomaly cancellation, poly-stability (supersymmetry), chirality of the spectrum, and the absence of exotic matter.

What carries the argument

LB-Explorer, a Transformer-based reinforcement learning environment that searches line bundle configurations on Calabi-Yau threefolds while satisfying multiple physical constraints through policy learning.

If this is right

  • Configurations discovered by the agent can be further filtered by equivariant structure of the line bundle sum and detailed particle spectrum requirements.
  • A hybrid setup with CP-SAT solvers can perturb RL solutions to impose selected conditions exactly.
  • The method extends to any Calabi-Yau threefold with a smooth simplicial Mori cone and freely acting discrete symmetry.
  • Such agents provide a tool for systematic exploration of the heterotic string landscape at large numbers of moduli.

Where Pith is reading between the lines

These are editorial extensions of the paper, not claims the author makes directly.

  • The learned constraint policies could be transferred to search problems on non-complete-intersection Calabi-Yau threefolds.
  • Integration with other machine learning or optimization techniques might enable automated generation of full particle spectra.
  • Similar reinforcement learning setups could be adapted to locate standard model vacua in different string theory constructions.

Load-bearing premise

The reinforcement learning architecture trained on specific Calabi-Yau examples will generalize its constraint satisfaction to any Calabi-Yau admitting a smooth, simplicial Mori cone and a freely-acting discrete symmetry.

What would settle it

Running LB-Explorer on a Calabi-Yau threefold outside the tested complete intersection set and checking whether all output configurations satisfy E8 embedding, anomaly cancellation, poly-stability, correct chirality, and absence of exotics without further manual adjustment.

Figures

Figures reproduced from arXiv: 2607.00078 by Alessandro Mininno, Gary Shiu, Jacky H. T. Yip.

Figure 1
Figure 1. Figure 1: Distribution of the maximal freely acting symmetry group orders across the favorable CICY [PITH_FULL_IMAGE:figures/full_fig_p010_1.png] view at source ↗
Figure 3
Figure 3. Figure 3: Schematic representation of the LB-Explorer architecture. [PITH_FULL_IMAGE:figures/full_fig_p014_3.png] view at source ↗
Figure 4
Figure 4. Figure 4: Average time of LB-Explorer for each seed. [PITH_FULL_IMAGE:figures/full_fig_p018_4.png] view at source ↗
Figure 5
Figure 5. Figure 5: Training metrics for three examples where learning clearly progresses in stages. [PITH_FULL_IMAGE:figures/full_fig_p019_5.png] view at source ↗
Figure 6
Figure 6. Figure 6: We show the accumulation of solutions for three examples of explorations. Figure [PITH_FULL_IMAGE:figures/full_fig_p020_6.png] view at source ↗
Figure 7
Figure 7. Figure 7: Number of total solutions found for CICY 6 over the ten seeds considered. [PITH_FULL_IMAGE:figures/full_fig_p021_7.png] view at source ↗
Figure 9
Figure 9. Figure 9: Weight distance and CKA heatmap for the transfer learning analysis from CICY 7447 to [PITH_FULL_IMAGE:figures/full_fig_p025_9.png] view at source ↗
Figure 11
Figure 11. Figure 11: Discovery Curve, CKA Heatmap and Policy Divergence for transfer learning analysis from [PITH_FULL_IMAGE:figures/full_fig_p026_11.png] view at source ↗
Figure 12
Figure 12. Figure 12: Discovery Curve, CKA Heatmap and Mean CRS for transfer learning analysis from CICY [PITH_FULL_IMAGE:figures/full_fig_p027_12.png] view at source ↗
Figure 13
Figure 13. Figure 13: Sample Efficiency and Weight distance for the transfer learning analysis from CICY 5302 [PITH_FULL_IMAGE:figures/full_fig_p028_13.png] view at source ↗
Figure 14
Figure 14. Figure 14: Sample Efficiency and Weight distance for the average transfer learning analysis from CICY [PITH_FULL_IMAGE:figures/full_fig_p028_14.png] view at source ↗
Figure 15
Figure 15. Figure 15: Discovery Curve and Mean CRS for transfer learning analysis from CICY 5302 ( [PITH_FULL_IMAGE:figures/full_fig_p029_15.png] view at source ↗
Figure 16
Figure 16. Figure 16: Discovery Curve and Mean CRS for transfer learning analysis from CICY 5452 ( [PITH_FULL_IMAGE:figures/full_fig_p030_16.png] view at source ↗
Figure 17
Figure 17. Figure 17: Discovery Curve and Policy Divergence for the transfer learning analysis from CICY 3413 [PITH_FULL_IMAGE:figures/full_fig_p031_17.png] view at source ↗
Figure 18
Figure 18. Figure 18: Sample Efficiency and Weight distance for the average transfer learning analysis from CICY [PITH_FULL_IMAGE:figures/full_fig_p031_18.png] view at source ↗
Figure 19
Figure 19. Figure 19: Discovery Curve, CKA Heatmap and Mean CRS for transfer learning analysis from CICY [PITH_FULL_IMAGE:figures/full_fig_p032_19.png] view at source ↗
Figure 20
Figure 20. Figure 20: Discovery Curve, CKA Heatmap and Mean CRS for transfer learning analysis from CICY [PITH_FULL_IMAGE:figures/full_fig_p033_20.png] view at source ↗
Figure 21
Figure 21. Figure 21: Percentage of solutions as in Table [PITH_FULL_IMAGE:figures/full_fig_p049_21.png] view at source ↗
Figure 22
Figure 22. Figure 22: Percentage of solutions as in Table [PITH_FULL_IMAGE:figures/full_fig_p059_22.png] view at source ↗
read the original abstract

We propose a Transformer-based Reinforcement Learning architecture, "LB-Explorer", to search for heterotic line bundle standard models arising from compactifications on smooth Calabi-Yau (CY) threefolds. We construct $E_8\times E_8$ vacua with $\text{SU}(5)$ symmetry, where the $\text{SU}(5)$ can be further broken to the Standard Model gauge group via discrete Wilson lines. We test the LB-Explorer environment on complete intersection Calabi-Yau (CICY) manifolds, though the neural network architecture naturally generalizes to any CY admitting a smooth, simplicial Mori cone and a freely-acting discrete symmetry. The LB-Explorer efficiently learns constraints on the line bundle sums, guaranteeing the $E_8$ gauge embedding, anomaly cancellation, poly-stability (supersymmetry), chirality of the spectrum, and the absence of exotic matter. Valid configurations can be subsequently filtered by imposing the missing constraints, such as the equivariant structure of the line bundle sum and further requirements on the particle spectrum. In this direction, we introduce a hybrid architecture incorporating CP-SAT solvers that aims to impose some of the conditions exactly by perturbing solutions found by the LB-Explorer. The versatility and scalability of the LB-Explorer make it a powerful tool for navigating the string landscape with a large number of moduli. The code and tools necessary to reproduce our findings are available at https://github.com/alexmininno/LB-Explorer

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, simulated authors' rebuttal, and a circularity audit. Tearing a paper down is the easy half of reading it; the pith above is the substance, this is the friction.

Referee Report

2 major / 0 minor

Summary. The manuscript proposes LB-Explorer, a Transformer-based reinforcement learning architecture to search for heterotic line bundle standard models on Calabi-Yau threefolds. It constructs E8 imes E8 vacua with SU(5) symmetry (broken to the Standard Model via discrete Wilson lines), tests the method on CICY manifolds, and asserts that the architecture naturally generalizes to any CY with a smooth simplicial Mori cone and freely-acting discrete symmetry. The approach is claimed to efficiently learn constraints guaranteeing E8 gauge embedding, anomaly cancellation, poly-stability, chirality, and absence of exotics, with a hybrid CP-SAT solver component for additional exact constraints; code is provided for reproducibility.

Significance. If the efficiency, constraint-learning guarantees, and generalization claims hold, the work would offer a scalable ML tool for navigating high-dimensional regions of the heterotic string landscape, complementing traditional algebraic geometry searches. The open-sourcing of code and tools is a clear strength that supports reproducibility.

major comments (2)
  1. [Abstract] Abstract: The assertion that 'the neural network architecture naturally generalizes to any CY admitting a smooth, simplicial Mori cone and a freely-acting discrete symmetry' is unsupported by transfer experiments, ablations on varying Mori cone generators or cohomology rings, or an argument establishing that the state representation (line-bundle integers plus intersection data) and reward formulation remain well-defined and trainable under changes to the discrete group action or cone structure. This extrapolation is load-bearing for the versatility and scalability claims.
  2. [Abstract] Abstract: No training curves, success rates, explicit formulations of the learned constraints, or validation against known line bundle models are visible, preventing assessment of whether the RL policy actually satisfies the claimed guarantees on E8 embedding, anomaly cancellation, poly-stability, chirality, and absence of exotics.

Simulated Author's Rebuttal

2 responses · 0 unresolved

We thank the referee for their careful reading and constructive comments. We address the major comments point by point below.

read point-by-point responses

  1. Referee: [Abstract] Abstract: The assertion that 'the neural network architecture naturally generalizes to any CY admitting a smooth, simplicial Mori cone and a freely-acting discrete symmetry' is unsupported by transfer experiments, ablations on varying Mori cone generators or cohomology rings, or an argument establishing that the state representation (line-bundle integers plus intersection data) and reward formulation remain well-defined and trainable under changes to the discrete group action or cone structure. This extrapolation is load-bearing for the versatility and scalability claims.

    Authors: The state representation is defined in terms of line-bundle integers together with the intersection numbers of the threefold, quantities that are well-defined for any smooth Calabi-Yau with a simplicial Mori cone. The reward function likewise encodes the physical constraints directly in these data. While the manuscript demonstrates the approach only on CICYs, the architecture itself does not encode manifold-specific features beyond these inputs. We agree that the generalization statement in the abstract is stated too strongly without supporting experiments and will revise it to a more qualified claim, adding a short discussion of the representation’s expected domain of applicability. revision: partial

  2. Referee: [Abstract] Abstract: No training curves, success rates, explicit formulations of the learned constraints, or validation against known line bundle models are visible, preventing assessment of whether the RL policy actually satisfies the claimed guarantees on E8 embedding, anomaly cancellation, poly-stability, chirality, and absence of exotics.

    Authors: The abstract is necessarily brief. Training curves, success rates, explicit constraint formulations, and comparisons to known line-bundle models appear in Sections 3 and 4 of the main text. To improve immediate visibility of these results we will add a concise performance summary to the abstract and ensure the constraint equations are stated explicitly in the revised introduction. revision: yes

Circularity Check

0 steps flagged

No circularity: search procedure with empirically tested constraints on CICYs

full rationale

The paper describes an RL-based search algorithm (LB-Explorer) whose reward is constructed to enforce the listed physical constraints (E8 embedding, anomaly cancellation, etc.). Finding valid configurations is therefore the intended output of the method rather than a derived prediction that reduces to its inputs. No mathematical derivation chain, self-citation load-bearing steps, or fitted-input-called-prediction patterns appear. The generalization claim to arbitrary CYs is an untested assertion but does not create circularity in any claimed result. The work is self-contained as a computational tool whose performance is evaluated directly on the manifolds used for training.

Axiom & Free-Parameter Ledger

0 free parameters · 0 axioms · 0 invented entities

Abstract-only review provides no explicit free parameters, axioms, or invented entities; the method relies on standard heterotic consistency conditions already present in the literature.

pith-pipeline@v0.9.1-grok · 5798 in / 1048 out tokens · 25209 ms · 2026-07-02T18:31:40.663616+00:00 · methodology

discussion (0)

Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.

Reference graph

Works this paper leans on

84 extracted references · 65 canonical work pages · 37 internal anchors

  1. [1]

    The Standard Model from String Theory: What Have We Learned?,

    F. Marchesano, G. Shiu, and T. Weigand, “The Standard Model from String Theory: What Have We Learned?,”Ann. Rev. Nucl. Part. Sci.74no. 1, (2024) 113–140,arXiv:2401.01939 [hep-th]

    work page arXiv 2024

  2. [2]

    Snowmass White Paper: String Theory and Particle Physics,

    M. Cvetic, J. Halverson, G. Shiu, and W. Taylor, “Snowmass White Paper: String Theory and Particle Physics,”arXiv:2204.01742 [hep-th]

    work page arXiv

  3. [3]

    Counting String Theory Standard Models

    A. Constantin, Y.-H. He, and A. Lukas, “Counting String Theory Standard Models,”Phys. Lett. B792(2019) 258–262,arXiv:1810.00444 [hep-th]

    work page internal anchor Pith review Pith/arXiv arXiv 2019

  4. [4]

    QuadrillionF-Theory Compactifications with the Exact Chiral Spectrum of the Standard Model,

    M. Cvetiˇ c, J. Halverson, L. Lin, M. Liu, and J. Tian, “QuadrillionF-Theory Compactifications with the Exact Chiral Spectrum of the Standard Model,”Phys. Rev. Lett.123no. 10, (2019) 101601,arXiv:1903.00009 [hep-th]

    work page arXiv 2019

  5. [5]

    The Heterotic String,

    D. J. Gross, J. A. Harvey, E. J. Martinec, and R. Rohm, “The Heterotic String,”Phys. Rev. Lett.54(1985) 502–505

    1985

  6. [6]

    Heterotic String Theory. 1. The Free Heterotic String,

    D. J. Gross, J. A. Harvey, E. J. Martinec, and R. Rohm, “Heterotic String Theory. 1. The Free Heterotic String,”Nucl. Phys. B256(1985) 253

    1985

  7. [7]

    Heterotic String Theory. 2. The Interacting Heterotic String,

    D. J. Gross, J. A. Harvey, E. J. Martinec, and R. Rohm, “Heterotic String Theory. 2. The Interacting Heterotic String,”Nucl. Phys. B267(1986) 75–124

    1986

  8. [8]

    Vacuum configurations for superstrings,

    P. Candelas, G. T. Horowitz, A. Strominger, and E. Witten, “Vacuum configurations for superstrings,”Nucl. Phys. B258(1985) 46–74

    1985

  9. [9]

    A Superstring Inspired Standard Model,

    B. R. Greene, K. H. Kirklin, P. J. Miron, and G. G. Ross, “A Superstring Inspired Standard Model,”Phys. Lett. B180(1986) 69

    1986

  10. [10]

    A Three Generation Superstring Model. 1. Compactification and Discrete Symmetries,

    B. R. Greene, K. H. Kirklin, P. J. Miron, and G. G. Ross, “A Three Generation Superstring Model. 1. Compactification and Discrete Symmetries,”Nucl. Phys. B278(1986) 667–693

    1986

  11. [11]

    A Three Generation Superstring Model. 2. Symmetry Breaking and the Low-Energy Theory,

    B. R. Greene, K. H. Kirklin, P. J. Miron, and G. G. Ross, “A Three Generation Superstring Model. 2. Symmetry Breaking and the Low-Energy Theory,”Nucl. Phys. B292(1987) 606–652

    1987

  12. [12]

    Two Hundred Heterotic Standard Models on Smooth Calabi-Yau Threefolds

    L. B. Anderson, J. Gray, A. Lukas, and E. Palti, “Two Hundred Heterotic Standard Models on Smooth Calabi-Yau Threefolds,”Phys. Rev. D84(2011) 106005,arXiv:1106.4804 [hep-th]

    work page internal anchor Pith review Pith/arXiv arXiv 2011

  13. [13]

    Heterotic Line Bundle Standard Models

    L. B. Anderson, J. Gray, A. Lukas, and E. Palti, “Heterotic Line Bundle Standard Models,” JHEP06(2012) 113,arXiv:1202.1757 [hep-th]

    work page internal anchor Pith review Pith/arXiv arXiv 2012

  14. [14]

    A Comprehensive Scan for Heterotic SU(5) GUT models

    L. B. Anderson, A. Constantin, J. Gray, A. Lukas, and E. Palti, “A Comprehensive Scan for Heterotic SU(5) GUT models,”JHEP01(2014) 047,arXiv:1307.4787 [hep-th]

    work page internal anchor Pith review Pith/arXiv arXiv 2014

  15. [15]

    Anomaly Cancellation in Supersymmetric D=10 Gauge Theory and Superstring Theory,

    M. B. Green and J. H. Schwarz, “Anomaly Cancellation in Supersymmetric D=10 Gauge Theory and Superstring Theory,”Phys. Lett. B149(1984) 117–122

    1984

  16. [16]

    Heterotic Model Building: 16 Special Manifolds

    Y.-H. He, S.-J. Lee, A. Lukas, and C. Sun, “Heterotic Model Building: 16 Special Manifolds,” JHEP06(2014) 077,arXiv:1309.0223 [hep-th]

    work page internal anchor Pith review Pith/arXiv arXiv 2014

  17. [17]

    CIPro Package: Complete Intersections in Products of Projective Spaces and Line Bundles

    L. B. Anderson, A. Constantin, J. Gray, Y.-H. He, S.-J. Lee, and A. Lukas, “CIPro Package: Complete Intersections in Products of Projective Spaces and Line Bundles,”arXiv:2606.27588 [hep-th]

    work page internal anchor Pith review Pith/arXiv arXiv

  18. [18]

    Explore and Exploit with Heterotic Line Bundle Models,

    M. Larfors and R. Schneider, “Explore and Exploit with Heterotic Line Bundle Models,” Fortsch. Phys.68no. 5, (2020) 2000034,arXiv:2003.04817 [hep-th]. 110

    work page arXiv 2020

  19. [19]

    Heterotic String Model Building with Monad Bundles and Reinforcement Learning,

    A. Constantin, T. R. Harvey, and A. Lukas, “Heterotic String Model Building with Monad Bundles and Reinforcement Learning,”Fortsch. Phys.70no. 2-3, (2022) 2100186, arXiv:2108.07316 [hep-th]

    work page arXiv 2022

  20. [20]

    Decoding Nature with Nature’s Tools: Heterotic Line Bundle Models of Particle Physics with Genetic Algorithms and Quantum Annealing,

    S. A. Abel, A. Constantin, T. R. Harvey, A. Lukas, and L. A. Nutricati, “Decoding Nature with Nature’s Tools: Heterotic Line Bundle Models of Particle Physics with Genetic Algorithms and Quantum Annealing,”Fortsch. Phys.72no. 2, (2024) 2300260,arXiv:2306.03147 [hep-th]

    work page arXiv 2024

  21. [21]

    Generating triangulations and fibrations with reinforcement learning,

    P. Berglund, G. Butbaia, Y.-H. He, E. Heyes, E. Hirst, and V. Jejjala, “Generating triangulations and fibrations with reinforcement learning,”Phys. Lett. B860(2025) 139158, arXiv:2405.21017 [hep-th]

    work page arXiv 2025

  22. [22]

    Attention Is All You Need

    A. Vaswani, N. Shazeer, N. Parmar, J. Uszkoreit, L. Jones, A. N. Gomez, L. Kaiser, and I. Polosukhin, “Attention Is All You Need,”arXiv e-prints(June, 2017) arXiv:1706.03762, arXiv:1706.03762 [cs.CL]

    work page internal anchor Pith review Pith/arXiv arXiv 2017

  23. [23]

    Complete Intersection Calabi-Yau Manifolds,

    P. Candelas, A. M. Dale, C. A. Lutken, and R. Schimmrigk, “Complete Intersection Calabi-Yau Manifolds,”Nucl. Phys. B298(1988) 493

    1988

  24. [24]

    Fibrations in CICY Threefolds

    L. B. Anderson, X. Gao, J. Gray, and S.-J. Lee, “Fibrations in CICY Threefolds,”JHEP10 (2017) 077,arXiv:1708.07907 [hep-th]

    work page internal anchor Pith review Pith/arXiv arXiv 2017

  25. [25]

    On Free Quotients of Complete Intersection Calabi-Yau Manifolds

    V. Braun, “On Free Quotients of Complete Intersection Calabi-Yau Manifolds,”JHEP04 (2011) 005,arXiv:1003.3235 [hep-th]

    work page internal anchor Pith review Pith/arXiv arXiv 2011

  26. [26]

    Discrete Symmetries of Complete Intersection Calabi-Yau Manifolds,

    A. Lukas and C. Mishra, “Discrete Symmetries of Complete Intersection Calabi-Yau Manifolds,”Commun. Math. Phys.379no. 3, (2020) 847–865,arXiv:1708.08943 [hep-th]

    work page arXiv 2020

  27. [27]

    Line bundle cohomologies on CICYs with Picard number two,

    M. Larfors and R. Schneider, “Line bundle cohomologies on CICYs with Picard number two,” Fortsch. Phys.67no. 12, (2019) 1900083,arXiv:1906.00392 [hep-th]

    work page arXiv 2019

  28. [28]

    Cohomology of Line Bundles: A Computational Algorithm

    R. Blumenhagen, B. Jurke, T. Rahn, and H. Roschy, “Cohomology of Line Bundles: A Computational Algorithm,”J. Math. Phys.51(2010) 103525,arXiv:1003.5217 [hep-th]

    work page internal anchor Pith review Pith/arXiv arXiv 2010

  29. [29]

    CYTools: A Software Package for Analyzing Calabi-Yau Manifolds,

    M. Demirtas, A. Rios-Tascon, and L. McAllister, “CYTools: A Software Package for Analyzing Calabi-Yau Manifolds,”arXiv:2211.03823 [hep-th]

    work page arXiv

  30. [30]

    Holes in Calabi-Yau Effective Cones,

    N. Gendler, E. Sheridan, M. Stillman, and D. H. Wu, “Holes in Calabi-Yau Effective Cones,” arXiv:2603.11173 [hep-th]

    work page arXiv

  31. [31]

    CP-SAT v9.12,

    L. Perron and F. Didier, “CP-SAT v9.12,” Google OR-Tools, 2025. https://developers.google.com/optimization/cp/cp_solver/

    2025

  32. [32]

    Monad Bundles in Heterotic String Compactifications

    L. B. Anderson, Y.-H. He, and A. Lukas, “Monad Bundles in Heterotic String Compactifications,”JHEP07(2008) 104,arXiv:0805.2875 [hep-th]

    work page internal anchor Pith review Pith/arXiv arXiv 2008

  33. [33]

    Heterotic Compactification, An Algorithmic Approach

    L. B. Anderson, Y.-H. He, and A. Lukas, “Heterotic Compactification, An Algorithmic Approach,”JHEP07(2007) 049,arXiv:hep-th/0702210

    work page internal anchor Pith review Pith/arXiv arXiv 2007

  34. [34]

    L. B. Anderson,Heterotic and M-theory Compactifications for String Phenomenology. Other thesis, University of Oxford, 8, 2008.arXiv:0808.3621 [hep-th]

    work page internal anchor Pith review Pith/arXiv arXiv 2008

  35. [35]

    Aspects of (2,0) String Compactifications,

    J. Distler and B. R. Greene, “Aspects of (2,0) String Compactifications,”Nucl. Phys. B304 (1988) 1–62

    1988

  36. [36]

    (0,2) Landau-Ginzburg Theory

    J. Distler and S. Kachru, “(0,2) Landau-Ginzburg theory,”Nucl. Phys. B413(1994) 213–243, arXiv:hep-th/9309110

    work page internal anchor Pith review Pith/arXiv arXiv 1994

  37. [37]

    Some Three Generation (0,2) Calabi-Yau Models

    S. Kachru, “Some three generation (0,2) Calabi-Yau models,”Phys. Lett. B349(1995) 76–82, arXiv:hep-th/9501131. 111

    work page internal anchor Pith review Pith/arXiv arXiv 1995

  38. [38]

    ANTI SELF-DUAL YANG-MILLS CONNECTIONS OVER COMPLEX ALGEBRAIC SURFACES AND STABLE VECTOR BUNDLES,

    S. K. Donaldson, “ANTI SELF-DUAL YANG-MILLS CONNECTIONS OVER COMPLEX ALGEBRAIC SURFACES AND STABLE VECTOR BUNDLES,”Proc. Lond. Math. Soc.50 (1985) 1–26

    1985

  39. [39]

    On the existence of hermitian-yang-mills connections in stable vector bundles,

    K. Uhlenbeck and S. T. Yau, “On the existence of hermitian-yang-mills connections in stable vector bundles,”Commun. Pure Appl. Math.39no. S1, (1986) S257–S293

    1986

  40. [40]

    Heterotic Models from Vector Bundles on Toric Calabi-Yau Manifolds

    Y.-H. He, S.-J. Lee, and A. Lukas, “Heterotic Models from Vector Bundles on Toric Calabi-Yau Manifolds,”JHEP05(2010) 071,arXiv:0911.0865 [hep-th]

    work page internal anchor Pith review Pith/arXiv arXiv 2010

  41. [41]

    Complete classification of reflexive polyhedra in four dimensions

    M. Kreuzer and H. Skarke, “Complete classification of reflexive polyhedra in four-dimensions,” Adv. Theor. Math. Phys.4(2000) 1209–1230,arXiv:hep-th/0002240

    work page internal anchor Pith review Pith/arXiv arXiv 2000

  42. [42]

    Stabilizing the Complex Structure in Heterotic Calabi-Yau Vacua

    L. B. Anderson, J. Gray, A. Lukas, and B. Ovrut, “Stabilizing the Complex Structure in Heterotic Calabi-Yau Vacua,”JHEP02(2011) 088,arXiv:1010.0255 [hep-th]

    work page internal anchor Pith review Pith/arXiv arXiv 2011

  43. [43]

    The Atiyah Class and Complex Structure Stabilization in Heterotic Calabi-Yau Compactifications

    L. B. Anderson, J. Gray, A. Lukas, and B. Ovrut, “The Atiyah Class and Complex Structure Stabilization in Heterotic Calabi-Yau Compactifications,”JHEP10(2011) 032, arXiv:1107.5076 [hep-th]

    work page internal anchor Pith review Pith/arXiv arXiv 2011

  44. [44]

    Vacuum Varieties, Holomorphic Bundles and Complex Structure Stabilization in Heterotic Theories

    L. B. Anderson, J. Gray, A. Lukas, and B. Ovrut, “Vacuum Varieties, Holomorphic Bundles and Complex Structure Stabilization in Heterotic Theories,”JHEP07(2013) 017, arXiv:1304.2704 [hep-th]

    work page internal anchor Pith review Pith/arXiv arXiv 2013

  45. [45]

    Formulae for Line Bundle Cohomology on Calabi-Yau Threefolds,

    A. Constantin and A. Lukas, “Formulae for Line Bundle Cohomology on Calabi-Yau Threefolds,”Fortsch. Phys.67no. 12, (2019) 1900084,arXiv:1808.09992 [hep-th]

    work page arXiv 2019

  46. [46]

    Machine Learning Line Bundle Cohomologies of Hypersurfaces in Toric Varieties

    D. Klaewer and L. Schlechter, “Machine Learning Line Bundle Cohomologies of Hypersurfaces in Toric Varieties,”Phys. Lett. B789(2019) 438–443,arXiv:1809.02547 [hep-th]

    work page internal anchor Pith review Pith/arXiv arXiv 2019

  47. [47]

    Machine Learning Line Bundle Cohomology,

    C. R. Brodie, A. Constantin, R. Deen, and A. Lukas, “Machine Learning Line Bundle Cohomology,”Fortsch. Phys.68no. 1, (2020) 1900087,arXiv:1906.08730 [hep-th]

    work page arXiv 2020

  48. [48]

    Geodesics in the extended K¨ ahler cone of Calabi-Yau threefolds,

    C. R. Brodie, A. Constantin, A. Lukas, and F. Ruehle, “Geodesics in the extended K¨ ahler cone of Calabi-Yau threefolds,”JHEP03(2022) 024,arXiv:2108.10323 [hep-th]

    work page arXiv 2022

  49. [49]

    Generating Functions for Line Bundle Cohomology Dimensions on Complex Projective Varieties,

    A. Constantin, “Generating Functions for Line Bundle Cohomology Dimensions on Complex Projective Varieties,”arXiv:2401.14463 [math.AG]

    work page arXiv

  50. [50]

    Gopakumar-Vafa hierarchies in winding inflation and uplifts,

    F. Carta, A. Mininno, N. Righi, and A. Westphal, “Gopakumar-Vafa hierarchies in winding inflation and uplifts,”JHEP05(2021) 271,arXiv:2101.07272 [hep-th]

    work page arXiv 2021

  51. [51]

    Kaleidoscopes, Waves and the Prepotential

    R. ´Alvarez-Garc´ ıa and F. Ruehle, “Kaleidoscopes, Waves and the Prepotential,” arXiv:2606.05280 [hep-th]

    work page internal anchor Pith review Pith/arXiv arXiv

  52. [52]

    COMPLETE INTERSECTION CALABI-YAU MANIFOLDS. 2. THREE GENERATION MANIFOLDS,

    P. Candelas, C. A. Lutken, and R. Schimmrigk, “COMPLETE INTERSECTION CALABI-YAU MANIFOLDS. 2. THREE GENERATION MANIFOLDS,”Nucl. Phys. B306(1988) 113

    1988

  53. [53]

    New Calabi-Yau Manifolds with Small Hodge Numbers

    P. Candelas and R. Davies, “New Calabi-Yau Manifolds with Small Hodge Numbers,”Fortsch. Phys.58(2010) 383–466,arXiv:0809.4681 [hep-th]

    work page internal anchor Pith review Pith/arXiv arXiv 2010

  54. [54]

    Completing the Web of $Z_3$ - Quotients of Complete Intersection Calabi-Yau Manifolds

    P. Candelas and A. Constantin, “Completing the Web ofZ 3 - Quotients of Complete Intersection Calabi-Yau Manifolds,”Fortsch. Phys.60(2012) 345–369,arXiv:1010.1878 [hep-th]

    work page internal anchor Pith review Pith/arXiv arXiv 2012

  55. [55]

    Rolling Among Calabi-Yau Vacua,

    P. Candelas, P. S. Green, and T. Hubsch, “Rolling Among Calabi-Yau Vacua,”Nucl. Phys. B 330(1990) 49

    1990

  56. [56]

    Moduli Space of Calabi-Yau Manifolds,

    P. Candelas and X. de la Ossa, “Moduli Space of Calabi-Yau Manifolds,”Nucl. Phys. B355 (1991) 455–481. 112

    1991

  57. [57]

    On the Connectedness of the Moduli Space of Calabi--Yau Manifolds

    A. C. Avram, P. Candelas, D. Jancic, and M. Mandelberg, “On the connectedness of moduli spaces of Calabi-Yau manifolds,”Nucl. Phys. B465(1996) 458–472,arXiv:hep-th/9511230

    work page internal anchor Pith review Pith/arXiv arXiv 1996

  58. [58]

    Triadophilia: A Special Corner in the Landscape

    P. Candelas, X. de la Ossa, Y.-H. He, and B. Szendroi, “Triadophilia: A Special Corner in the Landscape,”Adv. Theor. Math. Phys.12no. 2, (2008) 429–473,arXiv:0706.3134 [hep-th]

    work page internal anchor Pith review Pith/arXiv arXiv 2008

  59. [59]

    Proximal Policy Optimization Algorithms

    J. Schulman, F. Wolski, P. Dhariwal, A. Radford, and O. Klimov, “Proximal Policy Optimization Algorithms,”arXiv e-prints(July, 2017) arXiv:1707.06347,arXiv:1707.06347 [cs.LG]

    work page internal anchor Pith review Pith/arXiv arXiv 2017

  60. [60]

    Layer Normalization

    J. Lei Ba, J. R. Kiros, and G. E. Hinton, “Layer Normalization,”arXiv e-prints(July, 2016) arXiv:1607.06450,arXiv:1607.06450 [stat.ML]

    work page internal anchor Pith review Pith/arXiv arXiv 2016

  61. [61]

    Understanding the difficulty of training deep feedforward neural networks,

    X. Glorot and Y. Bengio, “Understanding the difficulty of training deep feedforward neural networks,” inProceedings of the Thirteenth International Conference on Artificial Intelligence and Statistics, Y. W. Teh and M. Titterington, eds., vol. 9 ofProceedings of Machine Learning Research, pp. 249–256. PMLR, Chia Laguna Resort, Sardinia, Italy, 13–15 may, ...

    2010

  62. [62]

    Reasonable Effectiveness of Random Weighting: A Litmus Test for Multi-Task Learning

    B. Lin, F. Ye, Y. Zhang, and I. W. Tsang, “Reasonable Effectiveness of Random Weighting: A Litmus Test for Multi-Task Learning.” 2022.https://arxiv.org/abs/2111.10603

    work page arXiv 2022

  63. [63]

    Exponential Moving Average of Weights in Deep Learning: Dynamics and Benefits

    D. Morales-Brotons, T. Vogels, and H. Hendrikx, “Exponential Moving Average of Weights in Deep Learning: Dynamics and Benefits.” 2024.https://arxiv.org/abs/2411.18704

    work page arXiv 2024

  64. [64]

    FAMO: Fast Adaptive Multitask Optimization

    B. Liu, Y. Feng, P. Stone, and Q. Liu, “FAMO: Fast Adaptive Multitask Optimization.” 2023. https://arxiv.org/abs/2306.03792

    work page arXiv 2023

  65. [65]

    Gradient Vaccine: Investigating and Improving Multi-task Optimization in Massively Multilingual Models

    Z. Wang, Y. Tsvetkov, O. Firat, and Y. Cao, “Gradient Vaccine: Investigating and Improving Multi-task Optimization in Massively Multilingual Models.” 2020. https://arxiv.org/abs/2010.05874

    work page arXiv 2020

  66. [66]

    High-Dimensional Continuous Control Using Generalized Advantage Estimation

    J. Schulman, P. Moritz, S. Levine, M. Jordan, and P. Abbeel, “High-Dimensional Continuous Control Using Generalized Advantage Estimation,”arXiv e-prints(June, 2015) arXiv:1506.02438,arXiv:1506.02438 [cs.LG]

    work page internal anchor Pith review Pith/arXiv arXiv 2015

  67. [67]

    Breeding Realistic D-Brane Models,

    G. J. Loges and G. Shiu, “Breeding Realistic D-Brane Models,”Fortsch. Phys.70no. 5, (2022) 2200038,arXiv:2112.08391 [hep-th]

    work page arXiv 2022

  68. [68]

    134 billion intersecting brane models,

    G. J. Loges and G. Shiu, “134 billion intersecting brane models,”JHEP12(2022) 097, arXiv:2206.03506 [hep-th]

    work page arXiv 2022

  69. [69]

    Vacua with Small Flux Superpotential,

    M. Demirtas, M. Kim, L. Mcallister, and J. Moritz, “Vacua with Small Flux Superpotential,” Phys. Rev. Lett.124no. 21, (2020) 211603,arXiv:1912.10047 [hep-th]

    work page arXiv 2020

  70. [70]

    Small Flux Superpotentials for Type IIB Flux Vacua Close to a Conifold,

    R. ´Alvarez-Garc´ ıa, R. Blumenhagen, M. Brinkmann, and L. Schlechter, “Small Flux Superpotentials for Type IIB Flux Vacua Close to a Conifold,”Fortsch. Phys.68(2020) 2000088,arXiv:2009.03325 [hep-th]

    work page arXiv 2020

  71. [71]

    Conifold Vacua with Small Flux Superpotential,

    M. Demirtas, M. Kim, L. McAllister, and J. Moritz, “Conifold Vacua with Small Flux Superpotential,”Fortsch. Phys.68(2020) 2000085,arXiv:2009.03312 [hep-th]

    work page arXiv 2020

  72. [72]

    Systematics of perturbatively flat flux vacua,

    F. Carta, A. Mininno, and P. Shukla, “Systematics of perturbatively flat flux vacua,”JHEP02 (2022) 205,arXiv:2112.13863 [hep-th]

    work page arXiv 2022

  73. [73]

    Systematics of perturbatively flat flux vacua for CICYs,

    F. Carta, A. Mininno, and P. Shukla, “Systematics of perturbatively flat flux vacua for CICYs,” JHEP08(2022) 297,arXiv:2201.10581 [hep-th]

    work page arXiv 2022

  74. [74]

    Toward Realistic Intersecting D-Brane Models

    R. Blumenhagen, M. Cvetic, P. Langacker, and G. Shiu, “Toward realistic intersecting D-brane models,”Ann. Rev. Nucl. Part. Sci.55(2005) 71–139,arXiv:hep-th/0502005. 113

    work page internal anchor Pith review Pith/arXiv arXiv 2005

  75. [75]

    Branes with Brains: Exploring String Vacua with Deep Reinforcement Learning

    J. Halverson, B. Nelson, and F. Ruehle, “Branes with Brains: Exploring String Vacua with Deep Reinforcement Learning,”JHEP06(2019) 003,arXiv:1903.11616 [hep-th]

    work page internal anchor Pith review Pith/arXiv arXiv 2019

  76. [76]

    Transforming Calabi-Yau Constructions: Generating New Calabi-Yau Manifolds with Transformers,

    J. H. T. Yip, C. Arnal, F. Charton, and G. Shiu, “Transforming Calabi-Yau Constructions: Generating New Calabi-Yau Manifolds with Transformers,”arXiv:2507.03732 [hep-th]

    work page arXiv

  77. [77]

    Generating Special Triangulations with Transformers

    C. Arnal, J. H. T. Yip, F. Charton, and G. Shiu, “Generating Special Triangulations with Transformers,” in . 6, 2026.arXiv:2606.26660 [hep-th]

    work page internal anchor Pith review Pith/arXiv arXiv 2026

  78. [78]

    https://www.gap-system.org

    The GAP Group,GAP – Groups, Algorithms, and Programming, Version 4.15.1, 2025. https://www.gap-system.org

    2025

  79. [79]

    https://www.sagemath.org

    The Sage Developers,SageMath, the Sage Mathematics Software System (Version 10.8), 2025. https://www.sagemath.org

    2025

  80. [80]

    Integral Cohomology and Mirror Symmetry for Calabi-Yau 3-folds

    V. Batyrev and M. Kreuzer, “Integral cohomology and mirror symmetry for Calabi-Yau 3-folds,”arXiv:math/0505432

    work page internal anchor Pith review Pith/arXiv arXiv

Showing first 80 references.