Pith. sign in

REVIEW 3 major objections 2 minor 44 references

A topological sum rule links a two-qubit gate’s geometric phases to the winding number of its driving Hamiltonian, and nontrivial topology is required for entanglement.

Reviewed by Pith at T0; open to challenge. T0 means a machine referee read the full paper against a public rubric. the ladder, T0–T4 →

Challenge this review Re-run · record.json Download PDF Read on arXiv ↗

T0 review · grok-4.5

2026-07-13 15:33 UTC pith:OV3ICZ7R

load-bearing objection Wrong paper in the cache: we only have the abstract of the topological sum-rule claim, so the result cannot be checked. the 3 major comments →

arxiv 2603.29795 v2 pith:OV3ICZ7R submitted 2026-03-31 quant-ph

Topological sum rule for geometric phases of quantum gates

Nadav Orion , Boris Rotstein , Nirron Miller , Eric Akkermans This is my paper
classification quant-ph PACS 03.65.Vf03.67.Lx03.65.Ud
keywords geometric phasetopological sum rulewinding numbertwo-qubit gatesWootters concurrenceentanglementquantum gates
verification ladder T0 review T1 audit T2 compute T3 formal T4 reserved

The pith

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

This paper claims that the geometric phases a two-qubit system picks up over a complete basis of initial states add up to a topological invariant of the driving Hamiltonian. The sum rule states that the total phase (in units of 2π) equals an integer multiple of the Hamiltonian’s winding number. Two implementations of the same gate that sit in different topological classes must therefore distribute those phases differently, and that difference can be read out with the Wootters concurrence. As a direct consequence, only Hamiltonians that can carry a nonzero winding number can generate entanglement: trivial topology is not enough.

Core claim

For two-qubit quantum gates the authors establish the identity ν_U = (1/2π) ∑_n γ_n = m ν_H: the sum of geometric phases over a complete basis of initial states equals an integer multiple of the winding number that classifies the driving Hamiltonian. Different topological classes of the same gate therefore produce different phase distributions, which are measurable via Wootters concurrence, and nontrivial topology (ν_H ≠ 0) is necessary for entanglement generation.

What carries the argument

The topological sum rule ν_U = (1/2π) ∑_n γ_n = m ν_H, which equates the total geometric phase of a complete basis to an integer multiple of the Hamiltonian winding number and thereby classifies gate implementations by topology.

Load-bearing premise

That a single winding number of the driving Hamiltonian fully classifies the relevant topology and that the complete-basis sum of geometric phases is well-defined and equals that winding number (up to the integer m) for the two-qubit gates under study.

What would settle it

Take two concrete two-qubit gate implementations that realize the same unitary but are driven by Hamiltonians of different winding numbers; measure the geometric phases over a complete initial-state basis (or the Wootters concurrence pattern). If the summed phases do not differ by an integer multiple of the winding-number difference, or if a ν_H = 0 drive still produces entanglement, the sum rule fails.

Watch this falsifier. Get emailed when new claim-graph text bears on it.

If this is right

  • Same gate, different topology: phase distributions must differ and can be distinguished by concurrence.
  • Entanglement generation requires access to nontrivial Hamiltonian topology (ν_H ≠ 0).
  • Topological class of a two-qubit gate becomes an experimentally accessible quantity via geometric-phase or concurrence measurements.
  • Gate design can treat winding number as a control resource for how geometric phases (and entanglement) are distributed.

Where Pith is reading between the lines

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

  • The same sum-rule logic may extend to multi-qubit or continuous-variable gates if a suitable winding number and complete-basis geometric-phase sum can be defined.
  • If concurrence is a faithful reporter of the phase distribution, tomography-light protocols could certify topological class without full process tomography.
  • Trivial-topology drives being unable to entangle suggests a topological obstruction that could constrain adiabatic or geometric gate libraries.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, simulated authors' rebuttal, and a circularity audit.

Referee Report

3 major / 2 minor

Summary. The submission is presented under the title and abstract of a quant-ph paper claiming a topological sum rule ν_U = (1/2π)∑_n γ_n = m ν_H for geometric phases of two-qubit gates, with a corollary that nontrivial Hamiltonian winding number ν_H is necessary for entanglement generation and that topological classes are distinguishable via Wootters concurrence. The supplied full manuscript body, however, is an unrelated eess.SP paper (internally labeled arXiv:2603.29796) proposing JEPA-MSAC, a joint-embedding predictive architecture for multimodal sensing-assisted mmWave V2I communications, with experiments on localization, beam prediction, and RSSI prediction using DeepSense 6G. No definitions of ν_U, ν_H, m, or γ_n, no derivation of the sum rule, and no quantum-gate or concurrence analysis appear in the body.

Significance. If the abstract’s topological sum rule and entanglement corollary were correctly derived and validated, the result would be of clear interest in geometric quantum computation and topological classification of two-qubit gates. That significance cannot be assessed from the present package: the load-bearing objects (winding number of the driving Hamiltonian, complete-basis sum of geometric phases, integer multiplicity m, and the concurrence-based distinction) are never defined or proved in the manuscript text that was provided. The communications content that is present is a separate, self-contained engineering contribution and does not support the quant-ph claims.

major comments (3)
  1. Title/abstract vs. full text: the abstract asserts ν_U = (1/2π)∑_n γ_n = m ν_H and an entanglement corollary for two-qubit gates, but §§I–V and all tables/figures develop JEPA-MSAC (multimodal tokenization, temporal block-masked JEPA pretraining, frozen-backbone localization/beam/RSSI heads). There is no section, equation, or experiment that defines ν_H, m, the basis sum of geometric phases, or Wootters concurrence. The central claim is therefore not present in the manuscript and cannot be refereed.
  2. Because the body is a different paper, the load-bearing conditions flagged in the abstract—domain of the sum rule, well-definedness of the complete-basis sum ∑_n γ_n, independence of m from the gate implementation, and necessity of ν_H ≠ 0 for entanglement—have no derivation, assumptions, or counterexample analysis to check. A quant-ph evaluation of soundness is impossible on the supplied text.
  3. Even reading the body on its own terms as a communications paper, it is not the work announced by the quant-ph title and abstract; accepting or revising under the stated claims would misrepresent the contribution. The package as submitted is not a coherent manuscript for the claimed result.
minor comments (2)
  1. Internal arXiv label in the body (2603.29796, eess.SP) disagrees with the header paper_id 2603.29795 (quant-ph), reinforcing a source/assembly error rather than a minor typesetting issue.
  2. If the JEPA-MSAC manuscript were submitted under its own title to an appropriate venue, presentation issues (notation rendering, figure captions, and baseline fairness) could be handled as ordinary revisions; they are irrelevant to the quant-ph claims under review here.

Circularity Check

0 steps flagged

No circularity found: supplied manuscript is an empirical JEPA-MSAC architecture paper with no first-principles derivation of the claimed topological sum rule.

full rationale

The CACHEABLE full text under the given paper_id is the unrelated JEPA-MSAC communications paper (internally arXiv:2603.29796), not a derivation of ν_U = (1/2π)∑_n γ_n = m ν_H. That manuscript proposes a self-supervised multimodal predictive backbone, freezes it, and trains lightweight heads for localization, beam prediction and RSSI; all claims are architectural or empirical, validated by ablations and baselines on DeepSense 6G. There are no self-definitional identities, parameters fitted then re-presented as predictions, load-bearing self-citations of uniqueness theorems, or renamed known results that force the reported metrics by construction. Because the topological sum-rule derivation chain is simply absent from the supplied text, no circular step can be exhibited by quotation or equation reduction. Score 0 is therefore the only warranted outcome under the hard rules.

Axiom & Free-Parameter Ledger

1 free parameters · 4 axioms · 1 invented entities

Only the abstract of the topological paper is available; the full text is a different paper. Ledger entries are therefore limited to quantities and assumptions named in the abstract. No free parameters can be fitted from data because no data or derivation for this claim is present. The winding number, geometric phases, and the integer m are taken as the abstract’s structural ingredients.

free parameters (1)
  • m (integer multiplicity in ν_U = m ν_H)
    Appears in the abstract sum rule without definition or derivation in the supplied text; may be fixed by topology or by Hilbert-space dimension, but cannot be checked here.
axioms (4)
  • domain assumption A two-qubit Hamiltonian is classified by a winding number ν_H that is an integer topological invariant of the control.
    Stated in the abstract as the quantity that classifies the Hamiltonian; standard in topological quantum control but not derived in the available text.
  • domain assumption Geometric phases γ_n accumulated over a complete basis of initial states are well-defined for the gate implementations considered.
    Required for the left-hand side of the sum rule; usual adiabatic or cyclic geometric-phase assumptions are not spelled out in the abstract.
  • domain assumption Wootters concurrence can distinguish different distributions of those geometric phases for the same logical gate.
    Abstract claims the distinction is measurable through concurrence; no proof or example is available here.
  • ad hoc to paper Nontrivial topology (ν_H ≠ 0) is necessary for the Hamiltonian to generate entanglement.
    Presented as a corollary of the sum rule; this is a strong claim that would require a precise definition of ‘generate entanglement’ and of the allowed Hamiltonians, neither of which is in the supplied text.
invented entities (1)
  • Topological sum rule ν_U = (1/2π) ∑_n γ_n = m ν_H no independent evidence
    purpose: Connects the total geometric phase over a complete basis to the Hamiltonian winding number and underpins the entanglement corollary.
    Central new relation claimed in the abstract; no independent derivation or external check is available in the provided material.

pith-pipeline@v1.1.0-grok45 · 22928 in / 2808 out tokens · 34613 ms · 2026-07-13T15:33:05.059342+00:00 · methodology

0 comments
read the original abstract

We establish a topological sum rule, $\nu_U = \frac{1}{2\pi}\sum_n\gamma_n = m\nu_H$, connecting the geometric phases accumulated by a two-qubit system over a complete basis of initial states to the winding number $\nu_H$ classifying its Hamiltonian. Implementations of the same gate from different topological classes must distribute these phases differently, making their distinction measurable through the Wootters concurrence. As a corollary, nontrivial topology is a necessary condition for entanglement: only Hamiltonians with access to $\nu_H \neq 0$ can generate it.

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

44 extracted references · 9 linked inside Pith

  1. [1]

    Study on artificial intelligence (AI)/machine learning (ML) for NR air interface (Release 19),

    3GPP, “Study on artificial intelligence (AI)/machine learning (ML) for NR air interface (Release 19),” 3rd Generation Partnership Project (3GPP), Technical Report (TR) 38.843, Sep. 2025, v19.0.0

    2025

  2. [2]

    AI-RAN in 6G networks: State-of-the- art and challenges,

    N. A. Khan and S. Schmid, “AI-RAN in 6G networks: State-of-the- art and challenges,”IEEE Open J. Commun. Soc., vol. 5, pp. 294–311, 2024

    2024

  3. [3]

    AI- RAN: Transforming RAN with AI-driven computing infrastructure,

    L. Kundu, X. Lin, R. Gadiyar, J.-F. Lacasse, and S. Chowdhury, “AI- RAN: Transforming RAN with AI-driven computing infrastructure,” IEEE Commun. Mag., vol. 64, no. 1, pp. 168–174, 2026

    2026

  4. [4]

    AI-Native 6G: Empowering intelligent RAN with accelerated compute,

    X. Lin, L. Kundu, S. Cammerer, Y . Huang, C. Dick, C. Santhosam, R. Gadiyar, R. Wiesmayr, and C. Studer, “AI-Native 6G: Empowering intelligent RAN with accelerated compute,”IEEE Wireless Commun., vol. 32, no. 6, pp. 11–14, 2025

    2025

  5. [5]

    An introduction to deep learning for the physical layer,

    T. O’Shea and J. Hoydis, “An introduction to deep learning for the physical layer,”IEEE Trans. Cogn. Commun. Netw., vol. 3, no. 4, pp. 563–575, 2017

    2017

  6. [6]

    Deep learning for massive MIMO CSI feedback,

    C.-K. Wen, W.-T. Shih, and S. Jin, “Deep learning for massive MIMO CSI feedback,”IEEE Wireless Commun. Lett., vol. 7, no. 5, pp. 748–751, 2018

    2018

  7. [7]

    ComNet: Combination of deep learning and expert knowledge in OFDM receivers,

    X. Gao, S. Jin, C.-K. Wen, and G. Y . Li, “ComNet: Combination of deep learning and expert knowledge in OFDM receivers,”IEEE Commun. Lett., vol. 22, no. 12, pp. 2627–2630, 2018

    2018

  8. [8]

    Deep learning-based channel estimation,

    M. Soltani, V . Pourahmadi, A. Mirzaei, and H. Sheikhzadeh, “Deep learning-based channel estimation,”IEEE Commun. Lett., vol. 23, no. 4, pp. 652–655, 2019

    2019

  9. [9]

    Large AI models for wireless physical layer,

    J. Guo, Y . Cui, S. Jin, and J. Zhang, “Large AI models for wireless physical layer,”IEEE Commun. Mag., pp. 1–8, 2026

    2026

  10. [10]

    LLM4CP: Adapting large language models for channel prediction,

    B. Liu, X. Liu, S. Gao, X. Cheng, and L. Yang, “LLM4CP: Adapting large language models for channel prediction,”J. Commun. Inf. Netw., vol. 9, no. 2, pp. 113–125, 2024

    2024

  11. [11]

    Beam prediction based on large language models,

    Y . Sheng, K. Huang, L. Liang, P. Liu, S. Jin, and G. Y . Li, “Beam prediction based on large language models,”IEEE Wireless Commun. Lett., pp. 1–5, 2025

    2025

  12. [12]

    BeamLLM: Vision- empowered mmWave beam prediction with large language models,

    C. Zheng, J. He, G. Cai, Z. Yu, and C. G. Kang, “BeamLLM: Vision- empowered mmWave beam prediction with large language models,” in Proc. IEEE 102nd Vehicular Technology Conf. (VTC2025-Fall), 2025, pp. 1–6

    2025

  13. [13]

    M2BeamLLM: Multimodal sensing-empowered mmWave beam pre- diction with large language models,

    C. Zheng, J. He, C. G. Kang, G. Cai, Z. Yu, and M. Debbah, “M2BeamLLM: Multimodal sensing-empowered mmWave beam pre- diction with large language models,”arXiv preprint arXiv:2506.14532, 2025

    Pith/arXiv arXiv 2025

  14. [14]

    WirelessGPT: A generative pre-trained multi-task learning framework for wireless communication,

    T. Yang, P. Zhang, M. Zheng, Y . Shi, L. Jing, J. Huang, and N. Li, “WirelessGPT: A generative pre-trained multi-task learning framework for wireless communication,”IEEE Network, vol. 39, no. 5, pp. 58–65, 2025

    2025

  15. [15]

    Large wireless model (LWM): A foundation model for wireless channels,

    S. Alikhani, G. Charan, and A. Alkhateeb, “Large wireless model (LWM): A foundation model for wireless channels,”arXiv preprint arXiv:2411.08872, 2024

    Pith/arXiv arXiv 2024

  16. [16]

    Large wireless localization model (LWLM): A foundation model for positioning in 6G networks,

    G. Pan, K. Huang, H. Chen, S. Zhang, C. H ¨ager, and H. Wymeersch, “Large wireless localization model (LWLM): A foundation model for positioning in 6G networks,”arXiv preprint arXiv:2505.10134, 2025

    Pith/arXiv arXiv 2025

  17. [17]

    LLM4WM: Adapting LLM for wireless multi-tasking,

    X. Liu, S. Gao, B. Liu, X. Cheng, and L. Yang, “LLM4WM: Adapting LLM for wireless multi-tasking,”IEEE Trans. Mach. Learn. Commun. Netw., vol. 3, pp. 835–847, 2025

    2025

  18. [18]

    Large language model enabled multi-task physical layer network,

    T. Zheng and L. Dai, “Large language model enabled multi-task physical layer network,”IEEE Trans. Commun., vol. 74, pp. 307–321, 2026

    2026

  19. [19]

    DeepSense 6G: A large-scale real-world multi-modal sensing and communication dataset,

    A. Alkhateeb, G. Charan, T. Osman, A. Hredzak, J. Morais, U. Demirhan, and N. Srinivas, “DeepSense 6G: A large-scale real-world multi-modal sensing and communication dataset,”IEEE Commun. Mag., vol. 61, no. 9, pp. 122–128, 2023

    2023

  20. [20]

    Vision-position multi-modal beam prediction using real millimeter wave datasets,

    G. Charan, T. Osman, A. Hredzak, N. Thawdar, and A. Alkhateeb, “Vision-position multi-modal beam prediction using real millimeter wave datasets,” inProc. IEEE Wireless Commun. Netw. Conf. (WCNC), 2022, pp. 2727–2731

    2022

  21. [21]

    Radar aided proactive blockage pre- diction in real-world millimeter wave systems,

    U. Demirhan and A. Alkhateeb, “Radar aided proactive blockage pre- diction in real-world millimeter wave systems,” inProc. IEEE Int. Conf. Commun. (ICC), 2022, pp. 4547–4552

    2022

  22. [22]

    LiDAR-aided mobile block- age prediction in real-world millimeter wave systems,

    S. Wu, C. Chakrabarti, and A. Alkhateeb, “LiDAR-aided mobile block- age prediction in real-world millimeter wave systems,” inProc. IEEE Wireless Commun. Netw. Conf. (WCNC), 2022, pp. 2631–2636

    2022

  23. [23]

    Position aided beam prediction in the real world: How useful GPS locations actually are?

    J. Morais, A. Behboodi, H. Pezeshki, and A. Alkhateeb, “Position aided beam prediction in the real world: How useful GPS locations actually are?” inProc. IEEE Int. Conf. Commun. (ICC), 2023, pp. 1824–1829

    2023

  24. [24]

    Deep learning for mmWave beam and blockage prediction using Sub-6 GHz channels,

    M. Alrabeiah and A. Alkhateeb, “Deep learning for mmWave beam and blockage prediction using Sub-6 GHz channels,”IEEE Trans. Commun., vol. 68, no. 9, pp. 5504–5518, 2020

    2020

  25. [25]

    Sensing- assisted high reliable communication: A transformer-based beamforming approach,

    Y . Cui, J. Nie, X. Cao, T. Yu, J. Zou, J. Mu, and X. Jing, “Sensing- assisted high reliable communication: A transformer-based beamforming approach,”IEEE J. Sel. Top. Signal Process., vol. 18, no. 5, pp. 782–795, 2024

    2024

  26. [26]

    Multimodal deep learning-empowered beam prediction in future THz ISAC systems,

    K. Zhang, W. Yu, H. He, S. Song, J. Zhang, and K. B. Letaief, “Multimodal deep learning-empowered beam prediction in future THz ISAC systems,” inProc. IEEE 36th Int. Symp. Pers. Indoor Mobile Radio Commun. (PIMRC), 2025, pp. 1–6

    2025

  27. [27]

    Language models meet world models: Embodied experiences enhance language models,

    J. Xiang, T. Tao, Y . Gu, T. Shu, Z. Wang, Z. Yang, and Z. Hu, “Language models meet world models: Embodied experiences enhance language models,” inProc. 37th Int. Conf. Neural Inf. Process. Syst. (NeurIPS). Curran Associates Inc., 2023, pp. 73 752–73 772

    2023

  28. [28]

    Large action models: From inception to implementation,

    L. Wang, F. Yang, C. Zhang, J. Lu, J. Qian, S. He, P. Zhao, B. Qiao, R. Huang, S. Qinet al., “Large action models: From inception to implementation,”arXiv preprint arXiv:2412.10047, 2024

    Pith/arXiv arXiv 2024

  29. [29]

    Embodied AI: From LLMs to world models,

    T. Feng, X. Wang, Y .-G. Jiang, and W. Zhu, “Embodied AI: From LLMs to world models,”arXiv preprint arXiv:2509.20021, 2025

    arXiv 2025

  30. [30]

    World models,

    D. Ha and J. Schmidhuber, “World models,”CoRR, vol. abs/1803.10122, 2018

    Pith/arXiv arXiv 2018

  31. [31]

    Recurrent world models facilitate policy evolution,

    ——, “Recurrent world models facilitate policy evolution,” inProc. 32nd Int. Conf. Neural Inf. Process. Syst.Red Hook, NY , USA: Curran Associates Inc., 2018, pp. 2455–2467

    2018

  32. [32]

    Self-supervised learning from images with a joint-embedding predictive architecture,

    M. Assran, Q. Duval, I. Misra, P. Bojanowski, P. Vincent, M. Rabbat, Y . LeCun, and N. Ballas, “Self-supervised learning from images with a joint-embedding predictive architecture,” inProc. IEEE/CVF Conf. Comput. Vis. Pattern Recognit. (CVPR), 2023, pp. 15 619–15 629

    2023

  33. [33]

    Revisiting feature prediction for learning visual representations from video,

    A. Bardes, Q. Garrido, J. Ponce, X. Chen, M. Rabbat, Y . LeCun, M. Assran, and N. Ballas, “Revisiting feature prediction for learning visual representations from video,”arXiv preprint arXiv:2404.08471, 2024

    Pith/arXiv arXiv 2024

  34. [34]

    V-JEPA 2: Self- supervised video models enable understanding, prediction and planning,

    M. Assran, A. Bardes, D. Fan, Q. Garrido, R. Howes, M. Muckley, A. Rizvi, C. Roberts, K. Sinha, A. Zholuset al., “V-JEPA 2: Self- supervised video models enable understanding, prediction and planning,” arXiv preprint arXiv:2506.09985, 2025

    Pith/arXiv arXiv 2025

  35. [35]

    Characterization of randomly time-variant linear channels,

    P. Bello, “Characterization of randomly time-variant linear channels,” IEEE Trans. Commun. Syst., vol. 11, no. 4, pp. 360–393, 1963

    1963

  36. [36]

    W. C. Jakes,Microwave Mobile Communications. New York: Wiley- IEEE Press, 1974

    1974

  37. [37]

    Finite-state Markov channel-a useful model for radio communication channels,

    H. S. Wang and N. Moayeri, “Finite-state Markov channel-a useful model for radio communication channels,”IEEE Trans. Veh. Technol., vol. 44, no. 1, pp. 163–171, 1995

    1995

  38. [38]

    Reading radio from camera: Visually-grounded, lightweight, and interpretable RSSI prediction,

    S. Yan, T. Hu, B. Mefgouda, S. Lasaulce, and M. Debbah, “Reading radio from camera: Visually-grounded, lightweight, and interpretable RSSI prediction,”arXiv preprint arXiv:2510.25936, 2025

    Pith/arXiv arXiv 2025

  39. [39]

    Vision aided channel prediction for vehicular communications: A case study of received power prediction using RGB images,

    X. Zhang, R. He, M. Yang, Z. Zhang, Z. Qi, and B. Ai, “Vision aided channel prediction for vehicular communications: A case study of received power prediction using RGB images,”IEEE Trans. Veh. Technol., vol. 74, no. 11, pp. 17 531–17 544, 2025

    2025

  40. [40]

    ImageNet large scale visual recognition challenge,

    O. Russakovsky, J. Deng, H. Su, J. Krause, S. Satheesh, S. Ma, Z. Huang, A. Karpathy, A. Khosla, M. Bernstein, A. C. Berg, and L. Fei- Fei, “ImageNet large scale visual recognition challenge,”Int. J. Comput. Vis., vol. 115, no. 3, pp. 211–252, 2015

    2015

  41. [41]

    EfficientNet: Rethinking model scaling for convolutional neural networks,

    M. Tan and Q. V . Le, “EfficientNet: Rethinking model scaling for convolutional neural networks,” inProc. 36th Int. Conf. Mach. Learn. (ICML), 2019, pp. 6105–6114

    2019

  42. [42]

    RankMe: Assessing the downstream performance of pretrained self-supervised representations by their rank,

    Q. Garrido, R. Balestriero, L. Najman, and Y . Lecun, “RankMe: Assessing the downstream performance of pretrained self-supervised representations by their rank,” inProc. 40th Int. Conf. Mach. Learn. (ICML), 2023, pp. 10 929–10 974

    2023

  43. [43]

    LiDAR: Sensing linear probing performance in joint embedding SSL architectures,

    V . Thilak, O. Saremi, P. Nakkiran, J. Susskind, C. Huang, H. Goh, L. Dinh, and E. Littwin, “LiDAR: Sensing linear probing performance in joint embedding SSL architectures,” inProc. ICLR/NeurIPS Workshop on SSL, 2023

    2023

  44. [44]

    Alpamayo-r1: Bridging reasoning and action prediction for generalizable autonomous driving in the long tail,

    Y . Wang, W. Luo, J. Bai, Y . Cao, T. Che, K. Chen, Y . Chen, J. Diamond, Y . Ding, W. Dinget al., “Alpamayo-r1: Bridging reasoning and action prediction for generalizable autonomous driving in the long tail,”arXiv preprint arXiv:2511.00088, 2025

    Pith/arXiv arXiv 2025