Pith. sign in

REVIEW 4 major objections 5 minor 31 references

CMU-Drive and V2V-VLA: Cooperative Multi-agent Unified Driving with Reasoning Benchmark and Vehicle-to-Vehicle Vision-Language-Action Models

T0 review · 4 major / 5 minor · reviewed 2026-08-11 · deepseek-v4-flash

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

Pith's one-line read Sharing a nearby car's view improves driving scores in simulation

desk verdict A genuinely useful cooperative driving benchmark held back by a confounded model comparison; referee should ask for an ablation. read the letter →

arxiv 2608.07621 v1 pith:FU5B3BDW submitted 2026-08-07 cs.AI cs.CVcs.RO

Hsu-kuang Chiu , Stephen F. Smith This is my paper · ORCID
classification cs.AIcs.CVcs.RO
keywords cooperativeautonomousdrivingvision-language-actionvehicle-to-vehiclecommunicationBEVoccupancymapclosed-loopbenchmarkmulti-agentlanguagereasoningend-to-end
verification ladder T0 review T1 audit T2 compute T3 formal

The pith

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

The reading

CMU-Drive is a closed-loop benchmark that extends single-agent urban driving scenarios to cooperative multi-agent settings, with 220 evaluation routes of 2 to 16 connected autonomous vehicles interacting with background vehicles, pedestrians, and cyclists across 44 safety-critical scenario types. The paper also proposes V2V-VLA, a vision-language-action model that, in one forward pass, generates driving actions, future waypoints, language reasoning, and a communication policy. V2V-VLA merges the ego vehicle's bird's-eye-view occupancy map with a selected neighboring vehicle's occupancy map, transformed into the ego frame, and uses this fused perception for closed-loop control. On CMU-Drive, V2V-VLA outperforms a non-cooperative VLA baseline, reaching a driving score of 63.67 and a success rate of 34.55 percent, versus 56.32 and 30.91 percent for the baseline. If correct, these results establish the first benchmark and baseline for cooperative vision-language-action driving and suggest that learned vehicle-to-vehicle communication can improve closed-loop multi-agent driving in safety-critical scenarios.

What carries the argument

The load-bearing mechanism is the merged bird's-eye-view occupancy map, $O_{\mathrm{merged}} = \mathrm{ElementwiseOr}(O_{\mathrm{ego}}, \mathrm{CoordinateTransform}(O_c, P_{\mathrm{ego}}, P_c))$, where $O_{\mathrm{ego}}$ and $O_c$ are camera-based BEV occupancy maps generated by UniAD, and the communicating vehicle's map is warped into the ego frame using both poses. The model also outputs a communication policy in language, recommending a different cooperative vehicle as the next communication target when the current one cannot see a critical object. These two pieces let the VLA model jointly reason about perception, action, and communication in a single forward pass, rather than relying on multi-round negotiation.

What would settle it

Run V2V-VLA on the 220 CMU-Drive routes with the communication channel disabled, for example by setting $O_{\mathrm{merged}}=O_{\mathrm{ego}}$ while keeping every other input identical; if the driving score and success rate do not drop below those of the baseline, the central claim that cooperative perception improves closed-loop driving would be refuted. A complementary check would retrain the model on the same data but without the merged occupancy map and compare the two.

Watch

Extended reading notes

Core claim

The central claim is that V2V-VLA achieves a superior driving score and a higher success rate than the non-cooperative baseline on the CMU-Drive benchmark. The paper argues that cooperative perception—merging the ego vehicle's BEV occupancy map $O_{\mathrm{ego}}$ with the coordinate-transformed occupancy map $O_{c\to\mathrm{ego}}$ of a chosen communicating vehicle via element-wise OR—plus an explicit communication policy that selects which neighbor to query next, is what delivers the improvement. This makes V2V-VLA the first cooperative vision-language-action driving model to be evaluated in closed loop on a multi-agent benchmark with up to 16 connected autonomous vehicles. The authors present the benchmark and this baseline as a foundation for future research on reasoning-guided, cooperative, end-to-end autonomous driving.

Load-bearing premise

The training data are collected by imitating PDM-lite, an expert that uses ground-truth location and velocity information for all objects in the simulation, whether visible or not, so the model's apparent advantage could come from imitating an oracle rather than from true cooperative reasoning.

Editorial extensions

If this is right

  • On CMU-Drive, switching from single-agent VLA driving to cooperative V2V-VLA raises driving score from 56.32 to 63.67 and success rate from 30.91 percent to 34.55 percent across all 220 routes.
  • The single-GPU evaluation protocol makes it practical to benchmark cooperative autonomous driving with 2 to 16 agents in closed loop, including scenarios with emergency vehicles and crossing pedestrians.
  • The model's language output provides interpretable explanations of when and why a neighbor is queried, which could support human oversight of cooperative decisions.
  • Because the benchmark and checkpoint will be released, future methods can be compared directly against the first cooperative VLA baseline without reimplementing the simulation harness.
  • The results suggest that fusing a neighbor's BEV occupancy, not just raw images, is sufficient to recover information about occluded critical objects and improve closed-loop safety.

Reading between the lines

Editorial extensions of the paper, not claims the author makes directly.

  • The expert used for data collection, PDM-lite, has access to ground-truth positions of all objects; the reported gap might therefore partly reflect the model imitating an oracle that sees through occlusions, rather than learning a generalizable cooperative skill.
  • A natural stress test would disable the communication channel entirely (e.g., set the merged map to the ego map) and measure the drop in driving score; the paper does not report an ablation of this kind.
  • The single-communicating-vehicle design could be extended to selecting among multiple views or to fusing temporal information, which might close the gap on the hardest scenario types reported (e.g., NonSignalizedJunctionLeftTurnEnterFlow).
  • Testing on CARLA routes where the communicating vehicle's camera is deliberately corrupted would verify that the model relies on the shared BEV map rather than spurious cues.
Share X Bluesky LinkedIn Reddit HN

Signed reviews

No signed human review yet.

Editorial analysis

A structured set of objections, weighed in public.

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

Referee Report

4 major / 5 minor

Summary. The paper introduces CMU-Drive, a closed-loop cooperative multi-agent driving benchmark that extends Bench2Drive by simulating 2 to 16 connected autonomous vehicles (CAVs) sharing the same CARLA environment, with 220 evaluation routes across 44 scenario types. The authors also propose V2V-VLA, a vision-language-action model that, in a single forward pass, generates driving waypoints, language reasoning, and a communication policy, using an ego camera image plus a merged BEV occupancy map from a selected neighboring CAV. The model is initialized from SimLingo and LoRA-fine-tuned on 180K training samples collected with the PDM-lite expert. The only quantitative comparison, reported in Table 2, shows V2V-VLA achieving a higher driving score (63.67 vs. 56.32) and success rate (34.55% vs. 30.91%) than the unmodified SimLingo baseline. The paper also includes one qualitative example illustrating a rain-blurred bicycle detected by a leading CAV via the proposed communication mechanism.

Significance. If fully validated, CMU-Drive would be a useful addition to the cooperative-driving benchmark landscape: it supports a substantially larger number of cooperating agents (up to 16) than InterDrive or MDrive, retains safety-critical scenario diversity from Bench2Drive, and claims single-GPU evaluation feasibility. The V2V-VLA model is a reasonable first baseline for this benchmark, and the authors commit to releasing code, benchmark, and checkpoints, which is a strength. However, the experimental support for the model's central claim is currently thin: the sole comparison in Table 2 is confounded, no ablations isolate the contribution of cooperation, and no statistical repeatability is reported. The benchmark itself would be more convincing with a breakdown by scenario type or by number of CAVs, and with a clearer demonstration that the proposed cooperative mechanism (not merely additional fine-tuning) is responsible for the reported gains.

major comments (4)
  1. [§4.1, Table 2] The comparison between V2V-VLA and SimLingo is confounded by at least three simultaneous differences: (i) V2V-VLA is LoRA-fine-tuned on 180K CMU-Drive samples distilled from the PDM-lite oracle, while SimLingo is used without any fine-tuning; (ii) V2V-VLA receives an extra merged BEV occupancy map from a neighboring CAV; and (iii) V2V-VLA is trained to emit a communication policy, while SimLingo is not. Because none of these factors is isolated, the observed DS/SR gains cannot be attributed to cooperation. Please add ablations, at minimum (a) SimLingo fine-tuned on the same 180K samples without the V2V branch, and (b) V2V-VLA with the merged-occupancy input removed or masked at inference. Without such a controlled comparison, the concluding claim in §5 that V2V-VLA achieves a superior driving score and success rate is not established.
  2. [§3.1 and §3.4] The learned communication policy is trained using annotations derived from ground-truth visibility and oracle knowledge of which CAV observes the critical object, yet the paper does not analyze how the policy behaves in closed-loop evaluation. Specifically, the paper does not report how often the model selects a non-default communicating CAV, whether those selections are correct, or what fraction of the driving-score improvement comes from the default closest-front-CAV communication versus the learned target switching. Since §3.1 states the model can override the default CAV, an evaluation of the communication decision accuracy and its direct effect on driving outcomes is needed to validate the reasoning-assisted communication component.
  3. [§2.2 and Table 2] All quantitative results are reported as single point estimates with no standard deviations, no confidence intervals, and no repeated evaluation seeds. In closed-loop CARLA evaluation, route initialization and nondeterministic traffic behavior can produce run-to-run variance, and the reported gaps (63.67 vs. 56.32 DS; 34.55% vs. 30.91% SR) may be within the noise range. Please either fix the random seeds and state this explicitly for all routes, or report aggregated results over multiple seeds so that the reader can judge the reliability of the difference.
  4. [§3.4] The training data are generated by PDM-lite, an expert that uses ground-truth locations and velocities of all objects 'regardless of whether objects are visible or invisible.' The communication supervision is therefore based on a perfect visibility model, whereas the deployed V2V-VLA perceives the world through its own camera and learned BEV occupancy. This creates a distribution shift: the model is trained to output communication requests when an object is invisible in ground truth, but at inference it must decide this from its own imperfect perception. The paper does not analyze how often the model's perceived visibility matches the ground-truth annotation, nor the consequences of misalignment for closed-loop safety. Please provide at least a quantitative comparison of ground-truth communication labels versus model-predicted communication targets on a held-out set, or explicitly discuss this limitation and its potential impact on the reported results.
minor comments (5)
  1. [§3.3] The section heading '3.3 Training' is followed immediately by '3.4 Training Data Collection' with no text under 3.3; this heading should either be removed or filled with a brief description of the training objective and procedure.
  2. [Table 1] The formatting in the CMU-Drive row of Table 1 is mangled: the entries for minimum, maximum, and average number of CAVs appear to be run together as '220 216 6.56 1'. Please present these values as separate columns with clear headers (Min. # CAVs, Max. # CAVs, Avg. # CAVs), as done for the other rows.
  3. [§3.2] In the text following Eq. (8), the word 'warpped' should be 'warped', and the notation for poses P_ego and P_c is used without a formal definition; please state that each pose includes position and orientation in a common coordinate frame.
  4. [Fig. 3] Figure 3 is mentioned in §4.1 as showing the average driving score per scenario type, but the figure is not displayed in the manuscript text provided. Please ensure all figures are included in the submission and that any color/legend information is legible.
  5. [References] Reference [1] for PDM-lite points to a personal GitHub report; if a formal publication or arXiv record exists, citing it would be more appropriate and verifiable.

Circularity Check

0 steps flagged · score 1.0 of 10

No significant circularity: the benchmark and model claims rest on external simulation metrics and supervised training, not on a self-referential derivation.

full rationale

The paper contains no derivation chain in which a predicted quantity is equal, by construction, to a fitted input. CMU-Drive is an evaluation benchmark that reuses Bench2Drive scenarios and the CARLA Leaderboard 2.0 metric (Eqs. 1-6), and V2V-VLA is a supervised model initialized from SimLingo and fine-tuned on 180K samples whose action labels come from PDM-lite, an oracle expert using ground-truth object states (Section 3.4). The communication-supervision annotations are generated from the same ground-truth simulation, but this is ordinary label construction for imitation learning; the closed-loop evaluation metric (DS/SR) is computed from route completion and infractions in CARLA, not from agreement with the training labels. The comparison in Table 2 is confounded—V2V-VLA receives extra fine-tuning on CMU-Drive data and an additional merged-occupancy input that SimLingo lacks, and no ablation isolates cooperation—but a confounded experiment is a validity concern, not circularity. Refs [3,4] are self-citations to the authors' prior V2V-LLM/V2V-GOT work, but they appear only in a related-work enumeration and are not load-bearing for the benchmark or model claims. One textual anomaly: Section 3.3 is titled 'Training' but contains no content; this is a missing-support issue that does not affect the circularity analysis. Overall, no step reduces to its own input by definition, so the paper is assessed as not significantly circular.

Assumptions & free parameters 0 free parameters · 4 assumptions · 0 invented entities

The benchmark formulas contain no fitted constants. The model itself is learned, and its training rests on domain assumptions about CARLA, PDM-lite, rule-based annotations, and the element-wise-OR fusion design; these are listed as axioms. No new physical or conceptual entities are postulated.

assumptions (4)
  • domain assumption Bench2Drive scenario definitions and safety-critical trigger logic are faithfully reproduced in CARLA.
    Section 2.1 inherits 44 scenario types from Bench2Drive; if these are not faithfully instantiated, the benchmark's difficulty and comparability claims break.
  • domain assumption PDM-lite is a competent expert for multi-agent data collection.
    Section 3.4 uses PDM-lite, which relies on ground-truth object states, as the source of action supervision for training; its choices directly shape the model's behavior.
  • domain assumption The rule-based language and communication annotations, similar to DriveLM and SimLingo, are correct enough for training.
    Section 3.4 generates language and communication targets with rules; incorrect labels would teach the model spurious reasoning or communication behavior.
  • ad hoc to paper Element-wise OR fusion of transformed BEV occupancy maps is a faithful representation of cooperative perception.
    Section 3.2 defines O_merged as ElementwiseOr(O_ego, O_c->ego); this design choice assumes no harmful occlusion, misalignment, or information loss in the fusion.

how reviews work

0 comments
Cite this review

Pith. "Pith review of CMU-Drive and V2V-VLA: Cooperative Multi-agent Unified Driving with Reasoning Benchmark and Vehicle-to-Vehicle Vision-Language-Action Models." pith.science (2026). https://pith.science/paper/FU5B3BDW

@misc{pith2026260807621,
  author       = {Pith},
  title        = {Pith review of: CMU-Drive and V2V-VLA: Cooperative Multi-agent Unified Driving with Reasoning Benchmark and Vehicle-to-Vehicle Vision-Language-Action Models},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/FU5B3BDW}},
  note         = {Machine review of arXiv:2608.07621}
}
read the original abstract

Vision-Language-Action (VLA) models have recently achieved impressive performance for end-to-end autonomous driving, yet existing approaches are primarily designed for an individual single autonomous driving agent with limited support for cooperative perception, reasoning, and planning. We present Cooperative Multi-agent Unified Driving with Reasoning (CMU-Drive), a closed-loop end-to-end benchmark for evaluating cooperative autonomous driving with multiple connected autonomous vehicles (CAVs) operating in safety-critical driving scenarios with background traffic participants. We further propose Vehicle-to-Vehicle Vision-Language-Action (V2V-VLA), a cooperative VLA model that integrates cooperative driving into a single forward pass by jointly generating driving actions, future waypoints, language reasoning, and communication policies. Experiments on CMU-Drive establish the first benchmark and baseline for cooperative VLA driving and provide a foundation for future research on multi-agent, closed-loop, end-to-end cooperative autonomous driving. Our code, benchmark, and model checkpoint will be publicly released to facilitate open-source research.

Figures

Figures reproduced from arXiv: 2608.07621 by the authors.

Figure 1
Figure 1. Evaluation route samples in CMU-Drive across different types of safety-critical scenarios and varying numbers of connected autonomous vehicles (CAVs). Each solid arrow represents a feasible path from the configured starting location to the destination of a CAV. Each dashed arrow represents the traffic flow of the configured background traffic participants within each safety-critical driving scenario. Each scenario a… view at source ↗
Figure 2
Figure 2. V2V-VLA model architecture. The input consists of a language prompt, tar￾get waypoints, and vision tokens derived from the ego vehicle’s front-camera image and a merged occupancy map from the ego and the selected communicating vehicle. The output comprises language outputs for the action, reasoning, and communication, alongside the future waypoints. 3 V2V-VLA: Vehicle-to-Vehicle Vision-Language-Action Model for Coop… view at source ↗
Figure 4
Figure 4. Qualitative experimental results. CAV0 is unable to observe the leading bicycle clearly due to the long distance and rain-induced image blur. Neverthe￾less, CAV0 still decides to reduce its speed by utilizing the perception information shared by the leading CAV4, which is closer to and capable of detecting the bi￾cycle. 4.2 Qualitative Result [PITH_FULL_IMAGE:figures/full_fig_p008_4.png] view at source ↗

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

31 extracted references · 16 canonical work pages

  1. [1]

    Beibwenger, J.: Pdm-lite: A rule-based planner for carla leaderboard 2.0 (2024), https://github.com/OpenDriveLab/DriveLM/blob/DriveLM- CARLA/pdm_lite/ docs/report.pdf

    work page 2024

  2. [2]

    In: Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition

    Chen, Z., Wu, J., Wang, W., Su, W., Chen, G., Xing, S., Zhong, M., Zhang, Q., Zhu, X., Lu, L., et al.: Internvl: Scaling up vision foundation models and aligning for generic visual-linguistic tasks. In: Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition. pp. 24185–24198 (2024)

    2024

  3. [3]

    IEEE International Conference on Robotics and Automation (ICRA) (2026)

    Chiu, H.k., Hachiuma, R., Wang, C.Y., Smith, S.F., Wang, Y.C.F., Chen, M.H.: V2v-llm: Vehicle-to-vehicle cooperative autonomous driving with multimodal large language models. IEEE International Conference on Robotics and Automation (ICRA) (2026)

    work page 2026

  4. [4]

    IEEE International Conference on Robotics and Automation (ICRA) (2026)

    Chiu, H.k., Hachiuma, R., Wang, C.Y., Wang, Y.C.F., Chen, M.H., Smith, S.F.: V2v-got: Vehicle-to-vehicle cooperative autonomous driving with multimodal large language models and graph-of-thoughts. IEEE International Conference on Robotics and Automation (ICRA) (2026)

    work page 2026

  5. [5]

    In: The Thirteenth International Conference on Learning Representations (2026)

    Cho, M., Cao, Y., Sun, J., Zhang, Q., Pavone, M., Park, J.J., Yang, H., Mao, Z.: Cocoon: Robust multi-modal perception with uncertainty-aware sensor fusion. In: The Thirteenth International Conference on Learning Representations (2026)

    work page 2026

  6. [6]

    MDrive: Benchmarking Closed-Loop Cooperative Driving for End-to-End Multi-agent Systems

    Coscoy, M., Zhou, Z., Zhao, S.Z., Wei, H., Magtoto, A., Liu, J., Song, R., Zimmer, W., Huang, Z., Tang, C., Zhou, B., Ma, J.: Mdrive: Benchmarking closed-loop cooperative driving for end-to-end multi-agent systems. arXiv preprint arXiv:2605.10904 (2026)

    work page Pith review arXiv 2026

  7. [7]

    In: Conference on Robot Learning (CoRL) (2017)

    Dosovitskiy, A., Ros, G., Codevilla, F., Lopez, A., Koltun, V.: CARLA: An open urban driving simulator. In: Conference on Robot Learning (CoRL) (2017)

    work page 2017

  8. [8]

    International Journal of Vehicular Technology2014(1), 259465 (2014).https://doi.org/https://doi

    Emirler, M.T., Uygan, I.M.C., Aksun Guvenc, B., Guvenc, L.: Robust pid steering control in parameter space for highly automated driving. International Journal of Vehicular Technology2014(1), 259465 (2014).https://doi.org/https://doi. org/10.1155/2014/259465,https://onlinelibrary.wiley.com/doi/abs/10. 1155/2014/259465

    work page doi:10.1155/2014/259465 2014

Show all 31 references
  1. [9]

    In: International Conference on Learning Representations (ICLR) (2022)

    Hu, E.J., Shen, Y., Wallis, P., Allen-Zhu, Z., Li, Y., Wang, S., Wang, L., Chen, W.: LoRA: Low-rank adaptation of large language models. In: International Conference on Learning Representations (ICLR) (2022)

    2022

  2. [10]

    arXiv preprint arXiv:2512.16760 (2025)

    Hu, T., Liu, X., Wang, S., Zhu, Y., Liang, A., Kong, L., Zhao, G., Gong, Z., Cen, J., Huang, Z., Hao, X., Li, L., Song, H., Li, X., Ma, J., Shen, S., Zhu, J., Tao, D., Liu, Z., Liang, J.: Vision-language-action models for autonomous driving: Past, present, and future. arXiv pr...

    2025

  3. [11]

    In: IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR) (2023)

    Hu,Y.,Yang,J.,Chen,L.,Li,K.,Sima,C.,Zhu,X.,Chai,S.,Du,S.,Lin,T.,Wang, W., et al.: Planning-oriented autonomous driving. In: IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR) (2023)

    2023

  4. [12]

    Hwang, J.J., Xu, R., Lin, H., Hung, W.C., Ji, J., Choi, K., Huang, D., He, T., Covington, P., Sapp, B., Zhou, Y., Guo, J., Anguelov, D., Tan, M.: Emma: End-to- endmultimodalmodelforautonomousdriving.In:arXivpreprintarXiv:2410.23262 (2024)

    2024 arXiv

  5. [13]

    In: NeurIPS 2024 Datasets and Benchmarks Track (2024)

    Jia, X., Yang, Z., Li, Q., Zhang, Z., Yan, J.: Bench2drive: Towards multi-ability benchmarking of closed-loop end-to-end autonomous driving. In: NeurIPS 2024 Datasets and Benchmarks Track (2024)

    2024

  6. [14]

    arXiv preprint arXiv:2512.10719 (2025) 10 H.-K

    Li, P., Zhang, Z., Holtz, D., Yu, H., Yang, Y., Lai, Y., Song, R., Geiger, A., Zell, A.: Spacedrive: Infusing spatial awareness into vlm-based autonomous driving. arXiv preprint arXiv:2512.10719 (2025) 10 H.-K. Chiu and S. F. Smith

    2025 arXiv

  7. [15]

    arXiv preprint arXiv:2503.08683 (2025)

    Liu, C., Liu, G., Wang, Z., Yang, J., Chen, S.: Colmdriver: Llm-based negotiation benefits cooperative autonomous driving. arXiv preprint arXiv:2503.08683 (2025)

    2025 arXiv

  8. [16]

    IEEE Transactions on Pattern Analysis and Machine Intelli- gence47(8), 6566–6584 (2025).https://doi.org/10.1109/TPAMI.2025.3560327

    Liu, G., Hu, Y., Xu, C., Mao, W., Ge, J., Huang, Z., Lu, Y., Xu, Y., Xia, J., Wang, Y., Chen, S.: Toward collaborative autonomous driving: Simulation platform and end-to-end system. IEEE Transactions on Pattern Analysis and Machine Intelli- gence47(8), 6566–6584 (2025).https:/...

    2025

  9. [17]

    arXiv preprint arXiv:2502.14156 (2025)

    Luo,K.Z.,Dao,M.Q.,Liu,Z.,Campbell,M.,Chao,W.L.,Weinberger,K.Q.,Malis, E., Fremont, V., Hariharan, B., Shan, M., Worrall, S., Berrio Perez, J.S.: Mixed signals: A diverse point cloud dataset for heterogeneous lidar v2x collaboration. arXiv preprint arXiv:2502.14156 (2025)

    2025 arXiv

  10. [18]

    In: Conference on Computer Vision and Pattern Recognition (CVPR) (2025)

    Renz, K., Chen, L., Arani, E., Sinavski, O.: Simlingo: Vision-only closed-loop au- tonomous driving with language-action alignment. In: Conference on Computer Vision and Pattern Recognition (CVPR) (2025)

    2025

  11. [19]

    In: Europian Conference on Computer Vision (ECCV) (2024)

    Sima, C., Renz, K., Chitta, K., Chen, L., Zhang, H., Xie, C., Luo, P., Geiger, A., Li, H.: Drivelm: Driving with graph visual question answering. In: Europian Conference on Computer Vision (ECCV) (2024)

    2024

  12. [20]

    arXiv preprint arXiv:2402.12289 (2024)

    Tian, X., Gu, J., Li, B., Liu, Y., Zhao, Z., Wang, Y., Zhan, K., Jia, P., Lang, X., Zhao, H.: Drivevlm: The convergence of autonomous driving and large vision- language models. arXiv preprint arXiv:2402.12289 (2024)

    2024 arXiv

  13. [21]

    arXiv preprint arXiv:2511.00088 (2025)

    Wang, L., Luo, H., et al.: Alpamayo-R1: Bridging reasoning and action pre- diction for generalizable autonomous driving in the long tail. arXiv preprint arXiv:2511.00088 (2025)

    2025 arXiv

  14. [22]

    In: European Conference on Computer Vision (ECCV) (2024)

    Xiang, H., Zheng, Z., Xia, X., Xu, R., Gao, L., Zhou, Z., Han, X., Ji, X., Li, M., Meng, Z., Jin, L., Lei, M., Ma, Z., He, Z., Ma, H., Yuan, Y., Zhao, Y., Ma, J.: V2x-real: a largs-scale dataset for vehicle-to-everything cooperative perception. In: European Conference on Compu...

    2024

  15. [23]

    arXiv (Dec 2024).https://doi.org/10.48550/arXiv.2412.15208

    Xing, S., Qian, C., Wang, Y., Hua, H., Tian, K., Zhou, Y., Tu, Z.: Openemma: Open-source multimodal model for end-to-end autonomous driving. arXiv (Dec 2024).https://doi.org/10.48550/arXiv.2412.15208

    2024 doi arXiv

  16. [24]

    In: IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR) (2023)

    Xu, R., Xia, X., Li, J., Li, H., Zhang, S., Tu, Z., Meng, Z., Xiang, H., Dong, X., Song, R., Yu, H., Zhou, B., Ma, J.: V2v4real: A real-world large-scale dataset for vehicle-to-vehicle cooperative perception. In: IEEE/CVF Conference on Computer Vision and Pattern Recognition (...

    2023

  17. [25]

    Yang, A., Yang, B., Hui, B., Zheng, B., Yu, B., Zhou, C., Li, C., Li, C., Liu, D., Huang, F., Dong, G., Wei, H., Lin, H., Tang, J., Wang, J., Yang, J., Tu, J., Zhang, J., Ma, J., Yang, J., Xu, J., Zhou, J., Bai, J., He, J., Lin, J., Dang, K., Lu, K., Chen, K., Yang, K., Li, M....

    2024 arXiv

  18. [26]

    In: 2025 IEEE/RSJ International Con- ference on Intelligent Robots and Systems (IROS)

    Zhao, S.Z., Xiang, H., Xu, C., Xia, X., Zhou, B., Ma, J.: Coopre: Cooperative pretraining for v2x cooperative perception. In: 2025 IEEE/RSJ International Con- ference on Intelligent Robots and Systems (IROS). pp. 11765–11772. IEEE (2025)

    2025

  19. [27]

    arXiv preprint arXiv:2509.03704 (2025)

    Zhao, S.Z., Zhang, H., Li, Z., Peng, J., Chui, A., Zhou, Z., Meng, Z., Xiang, H., Huang, Z., Wang, F., et al.: Quantv2x: A fully quantized multi-agent system for cooperative perception. arXiv preprint arXiv:2509.03704 (2025)

    2025 arXiv

  20. [28]

    arXiv preprint arXiv:2506.13757 (2025) CMU-Drive and V2V-VLA 11

    Zhou, Z., Cai, T., Zhao, Seth Z.and Zhang, Y., Huang, Z., Zhou, B., Ma, J.: Autovla: A vision-language-action model for end-to-end autonomous driving with adaptive reasoning and reinforcement fine-tuning. arXiv preprint arXiv:2506.13757 (2025) CMU-Drive and V2V-VLA 11

    2025 arXiv

  21. [29]

    arXiv preprint arXiv:2412.01812 (2024)

    Zhou, Z., Xiang, H., Zheng, Z., Zhao, S.Z., Lei, M., Zhang, Y., Cai, T., Liu, X., Liu, J., Bajji, M., Xia, X., Huang, Z., Zhou, B., Ma, J.: V2xpnp: Vehicle-to-everything spatio-temporal fusion for multi-agent perception and prediction. arXiv preprint arXiv:2412.01812 (2024)

    2024 arXiv

  22. [30]

    arXiv preprint arXiv:2508.04682 (2025)

    Zhou, Z., Zhao, S.Z., Cai, T., Huang, Z., Zhou, B., Ma, J.: Turbotrain: Towards ef- ficient and balanced multi-task learning for multi-agent perception and prediction. arXiv preprint arXiv:2508.04682 (2025)

    2025 arXiv

  23. [31]

    In: IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR) (2024)

    Zimmer, W., Wardana, G.A., Sritharan, S., Zhou, X., Song, R., Knoll, A.C.: Tum- traf v2x cooperative perception dataset. In: IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR) (2024)

    2024

Pith tools

Reviewed August 11, 2026 · model on record in the stance chip above.