pith. machine review for the scientific record. sign in

arxiv: 2605.08404 · v1 · submitted 2026-05-08 · 💻 cs.CL · cs.AI· cs.CV· cs.ET

Recognition: 2 theorem links

· Lean Theorem

Built Environment Reasoning from Remote Sensing Imagery Using Large Vision--Language Models

Dongdong Wang , Deepak Balakrishnan , Ravi Srinivasan , Shenhao Wang
Authors on Pith no claims yet

Pith reviewed 2026-05-12 01:05 UTC · model grok-4.3

classification 💻 cs.CL cs.AIcs.CVcs.ET
keywords remote sensing imagerylarge vision-language modelsbuilt environmentsmart citiesland use patternsrisk identificationconstructability assessment
0
0 comments X

The pith

Large vision-language models can generate built-environment recommendations from remote sensing images at varying scales.

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

This paper explores how large vision-language models process remote sensing imagery to reason about urban built environments. It tests imagery at different spatial scales as input for tasks like suggesting designs, assessing constructability, identifying land use patterns, and spotting risks. The work compares models such as InternVL and Qwen for their accuracy in producing reliable recommendations for smart city applications. A sympathetic reader would care because successful integration could enable data-driven urban planning and decision-making without needing extensive on-ground surveys. The results point to the feasibility of using these models to assist in complex city management tasks.

Core claim

The paper claims that remote sensing imagery at multiple spatial scales can be effectively used as input to large vision-language models for built-environment-related reasoning tasks, and that state-of-the-art models like InternVL and Qwen can produce accurate and reliable recommendations for design, constructability, landuse, and risk identification, demonstrating potential for smart cities.

What carries the argument

Multimodal language modeling with remote sensing imagery inputs at multiple scales, where the models generate text-based built environment recommendations and assessments.

If this is right

  • Integrating remote sensing with LLMs enables characterization of built environments for smart city tasks.
  • Different spatial scales affect the quality of reasoning outputs from the models.
  • Models like InternVL and Qwen show varying performance in accuracy and reliability for these tasks.
  • This approach can assist in decision-making for urban planning and risk management.

Where Pith is reading between the lines

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

  • If scaled, this could reduce the need for manual data collection in urban studies by leveraging existing satellite data.
  • Potential to extend to real-time monitoring if integrated with live satellite feeds.
  • Might connect to existing GIS systems for enhanced analytics in city management.

Load-bearing premise

The evaluation assumes that qualitative assessments of model outputs on remote sensing imagery will reliably indicate their usefulness for built-environment reasoning, even without detailed benchmarks or quantitative metrics.

What would settle it

A study that applies the models to a known urban area and measures how often their recommendations match actual expert analyses or on-site conditions would test the claim; if mismatch rate is high, the potential is overstated.

Figures

Figures reproduced from arXiv: 2605.08404 by Deepak Balakrishnan, Dongdong Wang, Ravi Srinivasan, Shenhao Wang.

Figure 1
Figure 1. Figure 1: Benchmark preparation pipeline. Built Environment Metrics We carefully select built environment metrics (Oliveira, 2020; Santhanam & Majumdar, 2022) to characterize spatial patterns of urban infrastructure, focusing on factors closely related to energy consumption management and large scale building development. These metrics capture both the structural composition and functional relationships of urban spa… view at source ↗
read the original abstract

This work investigates the use of large language models (LLMs) for tasks in smart cities. The core idea is to leverage remote sensing imagery to characterize the built environment, including design suggestions, constructability assessment, landuse patterns, and risk identification. We examine remote sensing imagery at multiple spatial scales as inputs for multimodal language modeling and evaluate their effects on built-environment-related reasoning. In addition, we compare state-of-the-art LLMs, including InternVL and Qwen, in terms of accuracy and reliability when generating built environment recommendations. The results demonstrate the potential of integrating remote sensing imagery with large language models to assist smart cities and decision-making.

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 / 2 minor

Summary. The manuscript investigates the use of large vision-language models (VLMs) such as InternVL and Qwen on remote sensing imagery for built-environment reasoning tasks in smart-city contexts. It processes imagery at multiple spatial scales to support tasks including design suggestions, constructability assessment, land-use pattern analysis, and risk identification, then compares the models on accuracy and reliability for generating recommendations. The central claim is that the results demonstrate the potential of integrating remote sensing imagery with VLMs to assist urban decision-making.

Significance. If the empirical comparisons were properly quantified and validated, the work could contribute to multimodal AI applications in urban informatics by showing how VLMs handle multi-scale remote sensing inputs for practical planning tasks. The absence of any reported metrics, benchmarks, or evaluation protocols, however, prevents assessment of whether the claimed potential is realized.

major comments (2)
  1. [Abstract] Abstract: the statement that 'the results demonstrate the potential' and that models were compared 'in terms of accuracy and reliability' is unsupported; no quantitative scores, benchmarks, datasets, annotation protocols, or validation methods are described, rendering the central empirical claim untestable.
  2. [Results] Results (or equivalent evaluation section): for open-ended reasoning outputs such as design suggestions and risk identification, accuracy is not self-evident; without an explicit evaluation setup, human judgment protocol, or inter-annotator agreement, the reliability comparison between InternVL and Qwen cannot be assessed.
minor comments (2)
  1. Clarify the exact spatial scales used and how they are encoded as input to the VLMs.
  2. Provide the full list of evaluation prompts or task templates employed for built-environment reasoning.

Simulated Author's Rebuttal

2 responses · 0 unresolved

We thank the referee for the insightful comments on our manuscript exploring large vision-language models for built environment reasoning from remote sensing imagery. We agree that the current presentation of results lacks sufficient detail on evaluation methods and overstates the empirical nature of the comparisons. We will revise the manuscript to clarify the qualitative nature of our analysis, add a dedicated evaluation section describing our assessment protocol, and temper the claims in the abstract and results to accurately reflect what was performed.

read point-by-point responses

  1. Referee: [Abstract] Abstract: the statement that 'the results demonstrate the potential' and that models were compared 'in terms of accuracy and reliability' is unsupported; no quantitative scores, benchmarks, datasets, annotation protocols, or validation methods are described, rendering the central empirical claim untestable.

    Authors: We agree that the abstract phrasing implies quantitative validation that is not supported by the manuscript. The comparisons between InternVL and Qwen were conducted qualitatively by examining the relevance, coherence, and practical utility of generated outputs for tasks like design suggestions and risk identification across multi-scale imagery. We will revise the abstract to state that the work presents exploratory case studies illustrating the potential of VLMs in this domain, without claiming quantitative accuracy or reliability metrics. We will also introduce a new evaluation subsection that specifies the imagery sources, the criteria applied during manual review (e.g., alignment with visible built environment features), and the limitations of the approach. revision: yes

  2. Referee: [Results] Results (or equivalent evaluation section): for open-ended reasoning outputs such as design suggestions and risk identification, accuracy is not self-evident; without an explicit evaluation setup, human judgment protocol, or inter-annotator agreement, the reliability comparison between InternVL and Qwen cannot be assessed.

    Authors: The referee correctly identifies that open-ended generative outputs require transparent evaluation protocols. Our original assessment involved author-led review of sample outputs for factual grounding in the remote sensing imagery, feasibility of recommendations, and consistency at different scales, but this was not formalized. In revision, we will expand the results section to detail the human judgment protocol, including the specific evaluation criteria used, the process for comparing model outputs, and acknowledgment that inter-annotator agreement was not quantified. We will also discuss the inherent difficulties in establishing quantitative benchmarks for these novel urban reasoning tasks and note this as a limitation of the current study. revision: yes

Circularity Check

0 steps flagged

No circularity: empirical comparison without derivations or self-referential steps

full rationale

The manuscript is an empirical study that feeds remote sensing imagery at multiple scales into VLMs (InternVL, Qwen) and reports qualitative comparisons on built-environment tasks. No equations, fitted parameters, predictions, or first-principles derivations appear in the provided text. The central claim—that the comparisons demonstrate potential for smart-city assistance—rests on the described model evaluations rather than any reduction to prior inputs by construction. No self-citations, ansatzes, or uniqueness theorems are invoked as load-bearing elements. The work is therefore self-contained against external benchmarks and receives the default non-circularity finding.

Axiom & Free-Parameter Ledger

0 free parameters · 1 axioms · 0 invented entities

The central claim depends on the unproven capability of current VLMs for this specialized reasoning task.

axioms (1)
  • domain assumption Multimodal LLMs can effectively reason about built environment characteristics from remote sensing imagery
    This is the core premise tested in the study.

pith-pipeline@v0.9.0 · 5413 in / 1201 out tokens · 58992 ms · 2026-05-12T01:05:40.888967+00:00 · methodology

discussion (0)

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

Lean theorems connected to this paper

Citations machine-checked in the Pith Canon. Every link opens the source theorem in the public Lean library.

What do these tags mean?
matches
The paper's claim is directly supported by a theorem in the formal canon.
supports
The theorem supports part of the paper's argument, but the paper may add assumptions or extra steps.
extends
The paper goes beyond the formal theorem; the theorem is a base layer rather than the whole result.
uses
The paper appears to rely on the theorem as machinery.
contradicts
The paper's claim conflicts with a theorem or certificate in the canon.
unclear
Pith found a possible connection, but the passage is too broad, indirect, or ambiguous to say the theorem truly supports the claim.

Reference graph

Works this paper leans on

295 extracted references · 295 canonical work pages · 2 internal anchors

  1. [1]

    Uncertainty quantification for traffic forecasting: A unified approach , pages =

    Qian, Weizhu and Zhang, Dalin and Zhao, Yan and Zheng, Kai and James,. Uncertainty quantification for traffic forecasting: A unified approach , pages =. 2023

    work page 2023

  2. [2]

    Proceedings of the web conference 2021 , author =

    work page 2021

  3. [3]

    A spatiotemporal analysis of the impact of lockdown and coronavirus on London’s bicycle hire scheme: from response to recovery to a new normal , pages =

    Gao, Xiaowei and Chen, Huanfa and Haworth, James , date =. A spatiotemporal analysis of the impact of lockdown and coronavirus on London’s bicycle hire scheme: from response to recovery to a new normal , pages =

    work page

  4. [4]

    Post-hoc uncertainty learning using a dirichlet meta-model , volume =

    Shen, Maohao and Bu, Yuheng and Sattigeri, Prasanna and Ghosh, Soumya and Das, Subhro and Wornell, Gregory , date =. Post-hoc uncertainty learning using a dirichlet meta-model , volume =. Proceedings of the

    work page

  5. [5]

    High-quality prediction intervals for deep learning: A distribution-free, ensembled approach , pages =

    Pearce, Tim and Brintrup, Alexandra and Zaki, Mohamed and Neely, Andy , date =. High-quality prediction intervals for deep learning: A distribution-free, ensembled approach , pages =. International conference on machine learning , publisher =

    work page

  6. [6]

    Probabilistic outputs for support vector machines and comparisons to regularized likelihood methods , volume =

    Platt, John and. Probabilistic outputs for support vector machines and comparisons to regularized likelihood methods , volume =

    work page

  7. [8]

    Temporal graph networks for deep learning on dynamic graphs , journaltitle =

    Rossi, Emanuele and Chamberlain, Ben and Frasca, Fabrizio and Eynard, Davide and Monti, Federico and Bronstein, Michael , date =. Temporal graph networks for deep learning on dynamic graphs , journaltitle =

    work page

  8. [9]

    Lower upper bound estimation method for construction of neural network-based prediction intervals , volume =

    Khosravi, Abbas and Nahavandi, Saeid and Creighton, Doug and Atiya, Amir F , date =. Lower upper bound estimation method for construction of neural network-based prediction intervals , volume =

    work page

  9. [10]

    Icml , author =

    Obtaining calibrated probability estimates from decision trees and naive bayesian classifiers , volume =. Icml , author =

    work page

  10. [11]

    Better uncertainty calibration via proper scores for classification and beyond , volume =

    Gruber, Sebastian and Buettner, Florian , date =. Better uncertainty calibration via proper scores for classification and beyond , volume =

    work page

  11. [12]

    Graph wavenet for deep spatial-temporal graph modeling , journaltitle =

    Wu, Zonghan and Pan, Shirui and Long, Guodong and Jiang, Jing and Zhang, Chengqi , date =. Graph wavenet for deep spatial-temporal graph modeling , journaltitle =

    work page

  12. [13]

    Self-attention convlstm for spatiotemporal prediction , volume =

    Lin, Zhihui and Li, Maomao and Zheng, Zhuobin and Cheng, Yangyang and Yuan, Chun , date =. Self-attention convlstm for spatiotemporal prediction , volume =. Proceedings of the

    work page

  13. [14]

    Accurate uncertainties for deep learning using calibrated regression , pages =

    Kuleshov, Volodymyr and Fenner, Nathan and Ermon, Stefano , date =. Accurate uncertainties for deep learning using calibrated regression , pages =. International conference on machine learning , publisher =

    work page

  14. [15]

    Proceedings of the 22nd international conference on Machine learning , author =

    Predicting good probabilities with supervised learning , pages =. Proceedings of the 22nd international conference on Machine learning , author =

    work page

  15. [16]

    Spatio-temporal graph convolutional networks: A deep learning framework for traffic forecasting , journaltitle =

    Yu, Bing and Yin, Haoteng and Zhu, Zhanxing , date =. Spatio-temporal graph convolutional networks: A deep learning framework for traffic forecasting , journaltitle =

    work page

  16. [17]

    Inductive graph neural networks for spatiotemporal kriging , volume =

    Wu, Yuankai and Zhuang, Dingyi and Labbe, Aurelie and Sun, Lijun , date =. Inductive graph neural networks for spatiotemporal kriging , volume =. Proceedings of the

    work page

  17. [18]

    Stop explaining black box machine learning models for high stakes decisions and use interpretable models instead , volume =

    Rudin, Cynthia , date =. Stop explaining black box machine learning models for high stakes decisions and use interpretable models instead , volume =

    work page

  18. [19]

    Uncertainty quantification in risk assessment-representation, propagation and treatment approaches: application to atmospheric dispersion modeling , volume =

    Abdo, H and Flaus, Jean-Marie and Masse, F , date =. Uncertainty quantification in risk assessment-representation, propagation and treatment approaches: application to atmospheric dispersion modeling , volume =

    work page

  19. [20]

    Adaptive modeling of uncertainties for traffic forecasting , journaltitle =

    Wu, Ying and Ye, Yongchao and Zeb, Adnan and Yu, James. Adaptive modeling of uncertainties for traffic forecasting , journaltitle =

    work page

  20. [21]

    Practical confidence and prediction intervals , journaltitle =

    Heskes, Tom , date =. Practical confidence and prediction intervals , journaltitle =

    work page

  21. [22]

    Risk analysis: a quantitative guide , publisher =

    Vose, David , date =. Risk analysis: a quantitative guide , publisher =

    work page

  22. [23]

    An algorithm for assessing the risk of traffic accident , volume =

    Ng, Kwok-suen and Hung, Wing-tat and Wong, Wing-gun , date =. An algorithm for assessing the risk of traffic accident , volume =

    work page

  23. [24]

    Building calibrated deep models via uncertainty matching with auxiliary interval predictors , volume =

    Thiagarajan, Jayaraman J and Venkatesh, Bindya and Sattigeri, Prasanna and Bremer, Peer-Timo , date =. Building calibrated deep models via uncertainty matching with auxiliary interval predictors , volume =. Proceedings of the

    work page

  24. [25]

    A survey of uncertainty in deep neural networks , journaltitle =

    Gawlikowski, Jakob and Tassi, Cedrique Rovile Njieutcheu and Ali, Mohsin and Lee, Jongseok and Humt, Matthias and Feng, Jianxiang and Kruspe, Anna and Triebel, Rudolph and Jung, Peter and Roscher, Ribana and. A survey of uncertainty in deep neural networks , journaltitle =

    work page

  25. [26]

    Quantifying uncertainty in deep spatiotemporal forecasting , pages =

    Wu, Dongxia and Gao, Liyao and Chinazzi, Matteo and Xiong, Xinyue and Vespignani, Alessandro and Ma, Yi-An and Yu, Rose , date =. Quantifying uncertainty in deep spatiotemporal forecasting , pages =. Proceedings of the 27th

    work page

  26. [27]

    Uncertainty quantification for traffic forecasting: a unified approach , journaltitle =

    Qian, Weizhu and Zhang, Dalin and Zhao, Yan and Zheng, Kai and Yu, James. Uncertainty quantification for traffic forecasting: a unified approach , journaltitle =

    work page

  27. [28]

    A bayesian learning network for traffic speed forecasting with uncertainty quantification , pages =

    Wu, Ying and James,. A bayesian learning network for traffic speed forecasting with uncertainty quantification , pages =. 2021 international joint conference on neural networks (

    work page 2021

  28. [29]

    Uncertainty quantification of spatiotemporal travel demand with probabilistic graph neural networks , journaltitle =

    Wang, Qingyi and Wang, Shenhao and Zhuang, Dingyi and Koutsopoulos, Haris and Zhao, Jinhua , date =. Uncertainty quantification of spatiotemporal travel demand with probabilistic graph neural networks , journaltitle =

    work page

  29. [30]

    Multi-type urban crime prediction , volume =

    Zhao, Xiangyu and Fan, Wenqi and Liu, Hui and Tang, Jiliang , date =. Multi-type urban crime prediction , volume =. Proceedings of the

    work page

  30. [31]

    Evaluating and calibrating uncertainty prediction in regression tasks , volume =

    Levi, Dan and Gispan, Liran and Giladi, Niv and Fetaya, Ethan , date =. Evaluating and calibrating uncertainty prediction in regression tasks , volume =

    work page

  31. [32]

    Measuring calibration in deep learning

    Nixon, Jeremy and Dusenberry, Michael W and Zhang, Linchuan and Jerfel, Ghassen and Tran, Dustin , date =. Measuring calibration in deep learning. , volume =

    work page

  32. [33]

    Distribution calibration for regression , pages =

    Song, Hao and Diethe, Tom and Kull, Meelis and Flach, Peter , date =. Distribution calibration for regression , pages =. International conference on machine learning , publisher =

    work page

  33. [34]

    Uncertainty quantification of sparse travel demand prediction with spatial-temporal graph neural networks , pages =

    Zhuang, Dingyi and Wang, Shenhao and Koutsopoulos, Haris and Zhao, Jinhua , date =. Uncertainty quantification of sparse travel demand prediction with spatial-temporal graph neural networks , pages =. Proceedings of the 28th

    work page

  34. [35]

    Streetgan: Towards road network synthesis with generative adversarial networks , url =

    Hartmann, Stefan and Weinmann, Michael and Wessel, Raoul and Klein, Reinhard , urldate =. Streetgan: Towards road network synthesis with generative adversarial networks , url =

    work page

  35. [36]

    Spatial planning of urban communities via deep reinforcement learning , volume =

    Zheng, Yu and Lin, Yuming and Zhao, Liang and Wu, Tinghai and Jin, Depeng and Li, Yong , date =. Spatial planning of urban communities via deep reinforcement learning , volume =

    work page

  36. [37]

    Neural Information Processing , publisher =

    Owaki, Takashi and Machida, Takashi , editor =. Neural Information Processing , publisher =. doi:10.1007/978-3-030-63820-7_61 , note =

    work page doi:10.1007/978-3-030-63820-7_61

  37. [39]

    Reimagining City Configuration: Automated Urban Planning via Adversarial Learning , isbn =

    Wang, Dongjie and Fu, Yanjie and Wang, Pengyang and Huang, Bo and Lu, Chang-Tien , urldate =. Reimagining City Configuration: Automated Urban Planning via Adversarial Learning , isbn =. Proceedings of the 28th International Conference on Advances in Geographic Information Systems , publisher =. doi:10.1145/3397536.3422268 , shorttitle =

    work page doi:10.1145/3397536.3422268

  38. [40]

    Quantifying the spatial homogeneity of urban road networks via graph neural networks , volume =

    Xue, Jiawei and Jiang, Nan and Liang, Senwei and Pang, Qiyuan and Yabe, Takahiro and Ukkusuri, Satish V and Ma, Jianzhu , date =. Quantifying the spatial homogeneity of urban road networks via graph neural networks , volume =

    work page

  39. [41]

    Neural turtle graphics for modeling city road layouts , url =

    Chu, Hang and Li, Daiqing and Acuna, David and Kar, Amlan and Shugrina, Maria and Wei, Xinkai and Liu, Ming-Yu and Torralba, Antonio and Fidler, Sanja , urldate =. Neural turtle graphics for modeling city road layouts , url =. Proceedings of the

    work page

  40. [42]

    Netgan: Generating graphs via random walks , url =

    Bojchevski, Aleksandar and Shchur, Oleksandr and Zügner, Daniel and Günnemann, Stephan , urldate =. Netgan: Generating graphs via random walks , url =. International conference on machine learning , publisher =

    work page

  41. [43]

    doi:10.1177/23998083211023516 , shorttitle =

    Ye, Xinyue and Du, Jiaxin and Ye, Yu , urldate =. doi:10.1177/23998083211023516 , shorttitle =

    work page doi:10.1177/23998083211023516

  42. [44]

    Machine learning assisted urban filling , url =

    Shen, Jiaqi and Liu, Chuan and Ren, Yue and Zheng, Hao , urldate =. Machine learning assisted urban filling , url =. Proceedings of the 25th

    work page

  43. [45]

    Human-instructed deep hierarchical generative learning for automated urban planning , volume =

    Wang, Dongjie and Wu, Lingfei and Zhang, Denghui and Zhou, Jingbo and Sun, Leilei and Fu, Yanjie , urldate =. Human-instructed deep hierarchical generative learning for automated urban planning , volume =. Proceedings of the

    work page

  44. [46]

    House-gan++: Generative adversarial layout refinement network towards intelligent computational agent for professional architects , url =

    Nauata, Nelson and Hosseini, Sepidehsadat and Chang, Kai-Hung and Chu, Hang and Cheng, Chin-Yi and Furukawa, Yasutaka , urldate =. House-gan++: Generative adversarial layout refinement network towards intelligent computational agent for professional architects , url =. Proceedings of the

    work page

  45. [47]

    doi:10.1080/13658816.2022.2041643 , shorttitle =

    Wu, Abraham Noah and Biljecki, Filip , urldate =. doi:10.1080/13658816.2022.2041643 , shorttitle =

    work page doi:10.1080/13658816.2022.2041643 2022

  46. [48]

    Development of an

    Park, Chulwoong and No, Wonjun and Choi, Junyong and Kim, Youngchul , urldate =. Development of an

    work page

  47. [49]

    Deep human-guided conditional variational generative modeling for automated urban planning , url =

    Wang, Dongjie and Liu, Kunpeng and Johnson, Pauline and Sun, Leilei and Du, Bowen and Fu, Yanjie , urldate =. Deep human-guided conditional variational generative modeling for automated urban planning , url =. 2021

    work page 2021

  48. [50]

    Building layout generation using site-embedded

    Jiang, Feifeng and Ma, Jun and Webster, Christopher John and Li, Xiao and Gan, Vincent. Building layout generation using site-embedded

    work page

  49. [51]

    Automated site planning using

    Jiang, Feifeng and Ma, Jun and Webster, Christopher John and Wang, Wei and Cheng, Jack. Automated site planning using

    work page

  50. [52]

    Understanding differences in residential location preferences between ownership and renting: A case study of London , volume =

    Haque, Md Bashirul and Choudhury, Charisma and Hess, Stephane , date =. Understanding differences in residential location preferences between ownership and renting: A case study of London , volume =

    work page

  51. [53]

    Housing location choices of the poor: does access to jobs matter? , volume =

    Hu, Lingqian and Wang, Liming , date =. Housing location choices of the poor: does access to jobs matter? , volume =

    work page

  52. [54]

    Modeling the choice of residential location , issn =

    work page

  53. [55]

    Understanding Dynamics of Travel Behavior with Inverse Reinforcement Learning and Hidden Markov Model , type =

    Yu, Mengqiao , date =. Understanding Dynamics of Travel Behavior with Inverse Reinforcement Learning and Hidden Markov Model , type =

    work page

  54. [56]

    Exogenous workplace choice in residential location models: is the assumption valid? , volume =

    Waddell, Paul , date =. Exogenous workplace choice in residential location models: is the assumption valid? , volume =

    work page

  55. [57]

    A deep gravity model for mobility flows generation , volume =

    Simini, Filippo and Barlacchi, Gianni and Luca, Massimilano and Pappalardo, Luca , date =. A deep gravity model for mobility flows generation , volume =

    work page

  56. [58]

    Mobility patterns are associated with experienced income segregation in large

    Moro, Esteban and Calacci, Dan and Dong, Xiaowen and Pentland, Alex , date =. Mobility patterns are associated with experienced income segregation in large

    work page

  57. [59]

    Network constraints on worker mobility , volume =

    Frank, Morgan R and Moro, Esteban and South, Tobin and Rutherford, Alex and Pentland, Alex and Taska, Bledi and Rahwan, Iyad , date =. Network constraints on worker mobility , volume =

    work page

  58. [60]

    Behavioral changes during the

    Yabe, Takahiro and Bueno, Bernardo García Bulle and Dong, Xiaowen and Pentland, Alex and Moro, Esteban , date =. Behavioral changes during the

    work page

  59. [61]

    Infrequent activities predict economic outcomes in major American cities , volume =

    Wang, Shenhao and Zheng, Yunhan and Wang, Guang and Yabe, Takahiro and Moro, Esteban and Pentland, Alex ‘Sandy’ , date =. Infrequent activities predict economic outcomes in major American cities , volume =

    work page

  60. [62]

    A deep inverse reinforcement learning approach to route choice modeling with context-dependent rewards , volume =

    Zhao, Zhan and Liang, Yuebing , urldate =. A deep inverse reinforcement learning approach to route choice modeling with context-dependent rewards , volume =

    work page

  61. [63]

    Learning Robust Rewards with Adversarial Inverse Reinforcement Learning , url =

    Fu, Justin and Luo, Katie and Levine, Sergey , urldate =. Learning Robust Rewards with Adversarial Inverse Reinforcement Learning , url =. 1710.11248 [cs] , keywords =

    work page arXiv

  62. [64]

    and Maas, Andrew L

    Ziebart, Brian D. and Maas, Andrew L. and Bagnell, J. Andrew and Dey, Anind K. , urldate =. Maximum entropy inverse reinforcement learning. , volume =. Aaai , publisher =

    work page

  63. [65]

    Choi, Seongjin and Kim, Jiwon and Yeo, Hwasoo , urldate =

    work page

  64. [66]

    Verified Uncertainty Calibration , url =

    Kumar, Ananya and Liang, Percy and Ma, Tengyu , urldate =. Verified Uncertainty Calibration , url =. 1909.10155 [cs, stat] , keywords =

    work page arXiv 1909

  65. [67]

    Uncertainty Quantification over Graph with Conformalized Graph Neural Networks , url =

    Huang, Kexin and Jin, Ying and Candes, Emmanuel and Leskovec, Jure , urldate =. Uncertainty Quantification over Graph with Conformalized Graph Neural Networks , url =. doi:10.48550/arXiv.2305.14535 , abstract =. 2305.14535 [cs, stat] , keywords =

    work page doi:10.48550/arxiv.2305.14535

  66. [68]

    Beyond Pinball Loss: Quantile Methods for Calibrated Uncertainty Quantification , url =

    Chung, Youngseog and Neiswanger, Willie and Char, Ian and Schneider, Jeff , urldate =. Beyond Pinball Loss: Quantile Methods for Calibrated Uncertainty Quantification , url =. doi:10.48550/arXiv.2011.09588 , shorttitle =. 2011.09588 [cs, stat] , keywords =

    work page doi:10.48550/arxiv.2011.09588 2011

  67. [69]

    Generative adversarial imitation learning , volume =

    Ho, Jonathan and Ermon, Stefano , urldate =. Generative adversarial imitation learning , volume =

    work page

  68. [70]

    Quantifying Uncertainty in Deep Spatiotemporal Forecasting , rights =

    Wu, Dongxia and Gao, Liyao and Xiong, Xinyue and Chinazzi, Matteo and Vespignani, Alessandro and Ma, Yi-An and Yu, Rose , urldate =. Quantifying Uncertainty in Deep Spatiotemporal Forecasting , rights =. doi:10.48550/ARXIV.2105.11982 , abstract =

    work page doi:10.48550/arxiv.2105.11982

  69. [71]

    PonderNet: Learning to ponder.arXiv preprint arXiv:2106.01345,

    Chen, Lili and Lu, Kevin and Rajeswaran, Aravind and Lee, Kimin and Grover, Aditya and Laskin, Michael and Abbeel, Pieter and Srinivas, Aravind and Mordatch, Igor , urldate =. Decision Transformer: Reinforcement Learning via Sequence Modeling , url =. doi:10.48550/arXiv.2106.01345 , shorttitle =

    work page doi:10.48550/arxiv.2106.01345

  70. [72]

    Uncertainty Quantification via Spatial-Temporal Tweedie Model for Zero-inflated and Long-tail Travel Demand Prediction , isbn =

    Jiang, Xinke and Zhuang, Dingyi and Zhang, Xianghui and Chen, Hao and Luo, Jiayuan and Gao, Xiaowei , urldate =. Uncertainty Quantification via Spatial-Temporal Tweedie Model for Zero-inflated and Long-tail Travel Demand Prediction , isbn =. Proceedings of the 32nd. doi:10.1145/3583780.3615215 , eventtitle =

    work page doi:10.1145/3583780.3615215

  71. [73]

    Uncertainty quantification of spatiotemporal travel demand with probabilistic graph neural networks , url =

    Wang, Qingyi and Wang, Shenhao and Zhuang, Dingyi and Koutsopoulos, Haris and Zhao, Jinhua , urldate =. Uncertainty quantification of spatiotemporal travel demand with probabilistic graph neural networks , url =

    work page

  72. [74]

    Vandal, E

    Lin, Kaixiang and Zhao, Renyu and Xu, Zhe and Zhou, Jiayu , urldate =. Efficient Large-Scale Fleet Management via Multi-Agent Deep Reinforcement Learning , isbn =. Proceedings of the 24th. doi:10.1145/3219819.3219993 , eventtitle =

    work page doi:10.1145/3219819.3219993

  73. [75]

    Combining revealed and stated preferences data , volume =

    Ben-Akiva, Moshe and Bradley, Mark and Morikawa, Takayuki and Benjamin, Julian and Novak, Thomas and Oppewal, Harmen and Rao, Vithala , date =. Combining revealed and stated preferences data , volume =

    work page

  74. [76]

    Conditional logit analysis of qualitative choice behavior , author =

    work page

  75. [77]

    Wang, Zepu and Zhuang, Dingyi and Li, Yankai and Zhao, Jinhua and Sun, Peng and Wang, Shenhao and Hu, Yulin , urldate =. 2023

    work page 2023

  76. [78]

    Modelling the choice of residential location , author =

    work page

  77. [79]

    Inductive graph neural networks for spatiotemporal kriging , volume =

    Wu, Yuankai and Zhuang, Dingyi and Labbe, Aurelie and Sun, Lijun , urldate =. Inductive graph neural networks for spatiotemporal kriging , volume =. Proceedings of the

    work page

  78. [81]

    Neural Combinatorial Optimization with Reinforcement Learning

    Bello, Irwan and Pham, Hieu and Le, Quoc V. and Norouzi, Mohammad and Bengio, Samy , urldate =. Neural Combinatorial Optimization with Reinforcement Learning , url =. doi:10.48550/arXiv.1611.09940 , abstract =

    work page Pith review doi:10.48550/arxiv.1611.09940

  79. [82]

    Attention, Learn to Solve Routing Problems! , url =

    Kool, Wouter and van Hoof, Herke and Welling, Max , urldate =. Attention, Learn to Solve Routing Problems! , url =

    work page

  80. [83]

    Multi-Agent Actor-Critic for Mixed Cooperative-Competitive Environments, March 2020

    Lowe, Ryan and Wu, Yi and Tamar, Aviv and Harb, Jean and Abbeel, Pieter and Mordatch, Igor , urldate =. Multi-Agent Actor-Critic for Mixed Cooperative-Competitive Environments , url =. doi:10.48550/arXiv.1706.02275 , abstract =. 1706.02275 [cs] , keywords =

    work page doi:10.48550/arxiv.1706.02275

Showing first 80 references.