Pith. sign in

REVIEW 3 major objections 5 minor 58 references

From Street Views to Urban Science: Discovering Road Safety Factors with Multimodal Large Language Models

T0 review · 3 major / 5 minor · reviewed 2026-08-07 · deepseek-v4-flash

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

Pith's one-line read The paper claims that an MLLM-driven loop of hypothesis generation, visual-question-answering embeddings, and significance-based pruning can discover interpretable road-safety factors that predict crash rates better than pretrained vision…

desk verdict Novel LLM-driven hypothesis loop with an honest case study, but the reported p-values are inflated by outcome-dependent selection, so the discovery claim needs a split-sample fix. read the letter →

arxiv 2506.02242 v2 pith:LZXQMWRU submitted 2025-06-02 cs.LG cs.CY

Yihong Tang , Ao Qu , Xujing Yu , Weipeng Deng , Jun Ma , Jinhua Zhao , Lijun Sun This is my paper
classification cs.LGcs.CY
keywords MultimodalLargeLanguageModelsStreetViewImageryInterpretableModelingHypothesisInferenceRoadSafetyCrashRatePredictionUrbanScienceVisualQuestionAnswering
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

This paper tries to establish that an automated pipeline can replace human hypothesis-making in urban science: a multimodal large language model (MLLM) proposes natural-language questions about street-view images, answers those questions, and feeds the answers as interpretable variables into regression models that predict crash rates. The framework, called UrbanX, prunes hypotheses whose coefficients are not significant and asks the language model to generate replacements, iterating until a compact set of transparent variables remains. On Manhattan road segments, the authors report that these interpretable embeddings predict crash rates better than fine-tuned ResNet and Vision Transformer baselines while preserving full traceability of each variable. If this holds, urban researchers would gain a scalable way to go from raw street imagery to statistically grounded, human-readable hypotheses about what makes streets safe.

What carries the argument

The carrying mechanism is the hypothesis-aligned embedding: each natural-language question is a hypothesis, and the MLLM's categorical answer to it is a dimension of the embedding vector for a street-view image. This makes every regression coefficient interpretable as the association between a concrete visual feature and crash rate. Around that embedding, Algorithm 1 runs an iterative posterior approximation: generate hypotheses from an LLM, extract embeddings via MLLM visual question answering, test each coefficient with a two-sided t-test, prune those with $p > \alpha$, and generate replacements conditioned on the surviving hypotheses and their p-values. The paper also uses SHAP values on a regression that combines the discovered variables with 58 conventional built-environment features to rank which factors matter most.

What would settle it

Refit the final 50 hypothesis variables on a held-out set of Manhattan segments and test whether the same coefficients stay significant; if they do not, the discovery claim is not supported. A second check is to recompute standard errors with spatial clustering by neighborhood and see whether the significant variables survive.

Watch

Extended reading notes

Core claim

UrbanX's central discovery is that visual question answering by an MLLM can turn unstructured street-view imagery into structured, interpretable variables that carry more predictive signal about segment-level crash rates than latent representations from pretrained deep learning models. The claim is operationalized as an iterative loop: a language model proposes hypotheses of the form 'Is there a median strip separating opposing traffic?', an MLLM answers them from each image, linear regression measures each variable's association with crash rates via t-tests, and insignificant hypotheses are pruned and replaced. The final 50 variables include median strips, lane markings, pedestrian crossings, visible pedestrians, and billboards; the paper reports that over 85 percent of variable pairs have absolute correlation below 0.2 and that the top contributors match established traffic-safety knowledge. The authors further claim the same loop can be transplanted to other urban outcomes such as walkability, equity, and environmental quality.

Load-bearing premise

The load-bearing premise is that the final p-values remain valid evidence even though the hypotheses were chosen using the same crash-rate data, and that regression errors are independent across adjacent road segments.

Editorial extensions

If this is right

  • An agency could train a crash-rate model whose inputs are plain-language questions about street views, so each prediction can be traced back to visible evidence.
  • The same generate-answer-test loop can be pointed at other street-view-linked outcomes, such as walkability, equity, or environmental quality, without manual variable engineering.
  • SHAP rankings of the discovered variables give planners a direct list of which visual features to change first, such as median strips, lane markings, or crossings.
  • On the Manhattan test set, the interpretable embeddings beat fine-tuned ResNet and Vision Transformer baselines on RMSE, MAE, and R-squared, indicating that transparency need not cost accuracy.

Reading between the lines

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

  • The paper computes p-values after using the same crash-rate outcome to prune hypotheses, so the reported significance levels are likely overstated; confirming the final set on independent data would settle how much of the discovery is real.
  • The regression assumes independent errors across adjacent Manhattan segments; if nearby streets share unmeasured risk factors, standard errors may be too small and some retained hypotheses might flip under spatially clustered inference.
  • A direct testable extension is to run the identical pipeline in a second city; hypotheses that replicate across cities would be far stronger evidence of generalizable safety factors.
  • The manual audit blames most MLLM errors on image resolution rather than reasoning, which predicts that higher-resolution or zoomed imagery should improve the long tail of hypotheses with little architectural change.
Share X Bluesky LinkedIn Reddit HN

Editorial analysis

A structured set of objections, weighed in public.

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

Referee Report

3 major / 5 minor

Summary. The paper proposes UrbanX, a framework that uses multimodal large language models (MLLMs) to generate natural-language hypotheses about visual features in street view images, converts the hypotheses into categorical embeddings via visual question answering, and evaluates them with linear regression in an iterative loop. The framework is applied to segment-level crash rates in Manhattan. The authors claim that UrbanX outperforms pretrained ResNet and ViT baselines while providing interpretable variables, and that the retained hypotheses are statistically significant correlates of crash rates. The central methodological contribution is an automated, LLM-driven hypothesis generation and refinement pipeline that is intended to support transparent, data-driven urban science.

Significance. If the claims were validated, UrbanX would be a useful template for transforming unstructured street-level imagery into interpretable, testable variables for urban safety research, and the idea of using MLLMs as semantic engines in an iterative statistical loop is timely and interesting. The authors provide a public code repository and a manual audit of MLLM answers, which are helpful for reproducibility. However, the current empirical evidence for both main claims—predictive superiority and statistically significant discovery—is undermined by statistical issues in the evaluation protocol. The paper's contribution is therefore promising but not yet established; the framework deserves publication only after the inferential problems are addressed.

major comments (3)
  1. [§3.4, Eq. (4)] The regression model in Eq. (4) assumes independent, normally distributed errors, but crash rates on spatially adjacent Manhattan road segments are likely spatially autocorrelated. The reported standard errors and p-values are therefore probably downward-biased, and the significance of the discovered factors is overstated. The authors should use spatially clustered standard errors, a spatial error model, or another approach that accounts for spatial dependence, and report how the results change.
  2. [§4.2, Figure 3] The claim that UrbanX 'outperforms pretrained deep learning models' is not supported empirically because Figure 3 reports only point estimates without error bars, confidence intervals, or significance tests. It is not possible to determine whether the differences between UrbanX (LR/LGBM) and the ResNet/ViT baselines are meaningful or due to random variation across data splits or model initializations. The authors should provide multiple runs or cross-validation folds with uncertainty estimates and a statistical comparison (e.g., paired tests) before claiming superiority.
  3. [§4.5, Figure 7] The robustness analysis reports 'convergence over 50 training epochs' for what appear to be linear regression and LightGBM models; linear regression has no epochs and LightGBM has boosting iterations, not epochs. This inconsistency makes it difficult to interpret the convergence curves. The authors should clarify the training procedure and the x-axis units.
minor comments (5)
  1. [Abstract and §2] The phrase 'full interpretability' is stronger than what is demonstrated: the framework uses SHAP values, which are approximate, and the final model includes both traditional features and hypothesis embeddings. Please temper the claim to 'interpretable' or specify the sense in which interpretability is full.
  2. [§4.4] The pairwise correlation analysis reports that over 85% of variable pairs have absolute correlation below 0.2, but Appendix C acknowledges semantic overlap among several hypotheses (e.g., the multi-lane and road-width questions, and the two barrier questions). Clarify the relationship between low linear correlation and semantic redundancy.
  3. [§4.6] The manual audit is performed on a single panoramic SVI and a separate random sample of 50 SVIs, but the secondary inspection results are not quantified in the table. Providing a confusion matrix or summary statistics for the 50-SVI audit would strengthen the reliability claim.
  4. [Figure 3] The y-axis label 'R20.0' appears to be a rendering artifact; the correct label should be 'R²'. Please fix the figure.
  5. [§4.7 and Appendix C] The limitations section acknowledges MLLM reliability and computational overhead, but it does not mention the selection-bias issue or spatial autocorrelation that affect the statistical conclusions. The appendix already notes redundancy and ambiguity in the hypothesis set; these concerns should be integrated into the main limitations discussion.

Circularity Check

1 steps flagged · score 6.0 of 10

Algorithm 1 selects hypotheses using p-values from the same crash-rate regression that later produces the reported significance evidence in §4.4.

  1. fitted input called prediction [Section 3.4 (hypothesis assessment, Eq. 4–5), Section 3.5 (Algorithm 1), Section 4.4 (Figure 6)]
    "We then apply a two-sided t-test to each coefficient β_j to assess the null hypothesis that β_j = 0 ... This yields a p-value p_t_j ... Hypotheses with p_t_j > α (typically α = 0.05) are considered statistically insignificant and are pruned in the next round. ... The left panel of Figure 6 visualizes each hypothesis according to its average SHAP value (x-axis), reflecting predictive contribution, and the negative base-10 logarithm of its p-value from linear regression (y-axis), reflecting statistical significance."

    The final 'significant' hypotheses displayed in Figure 6 are the survivors of an iterative loop that prunes hypotheses whose regression p-values exceed 0.05, using the same crash-rate outcome and the same linear model (Eq. 4–5, Algorithm 1). Under the null, variables that happen to achieve small p-values on the training or validation data are preferentially retained, so the distribution of p-values for the retained set is no longer uniform. The paper then reports these p-values as if they were unconditional evidence of discovery, without sample splitting, data carving, selective-inference correction, or any multiple-testing adjustment across the iterative rounds.

full rationale

The reported -log10(p) values in §4.4 are not independent evidence: Algorithm 1 progressively discards hypotheses with p > 0.05 and the final retained set is therefore enriched for small p-values from the same crash-rate regression that later supplies the significance plot. This is a statistical circularity built into the discovery workflow, and it directly affects the claim of discovering statistically meaningful road-safety factors. The predictive-performance comparison against ResNet and ViT is a separate, externally benchmarked claim and is not circular; the SHAP analysis and the manual MLLM audit also provide some independent content. Self-citations (Tang et al. 2024, 2025; Xia et al. 2025) appear only as related-work context and are not load-bearing justifications, so they do not raise the score. The spatial-independence assumption in Eq. 4 is a validity concern about downward-biased standard errors but is not itself a circularity; it would further weaken the p-value evidence if addressed. Overall, one central 'discovery' step reduces, by the paper's own selection rule, to selecting on the outcome it later claims to have found significant, warranting a partial-circularity score of 6 rather than a higher score.

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

The central claim rests on hand-chosen hyperparameters (p_explore=0.1, alpha=0.05, k=50), on the assumption that MLLM answers are reliable measurements, and on standard but violated statistical assumptions (i.i.d. errors, fixed hypothesis set). No new physical entities are introduced.

free parameters (4)
  • p_explore = 0.1
    Probability of using the exploration prompt in hypothesis generation; chosen by hand in Appendix A.1.
  • alpha = 0.05
    Significance threshold for pruning hypotheses; chosen by hand in Sections 3.1 and 3.4.
  • k (number of hypotheses) = 50
    Set based on validation performance; Section 4.5 shows optimal performance around 50 hypotheses.
  • T (number of iterations) = not specified
    The number of iterative refinement rounds is not clearly reported, yet it determines the final hypothesis set.
assumptions (4)
  • domain assumption Errors in the linear model (Eq. 4) are independent and normally distributed.
    Road segments in Manhattan are spatially correlated, yet the model uses ordinary least squares with no spatial or clustered errors, so standard errors are likely underestimated.
  • domain assumption MLLM categorical answers to generated hypotheses are reliable measurements of visual features.
    The framework treats VQA answers as ground-truth-like variables; evidence is limited to a one-image audit and a 50-SVI random check in Section 4.6.
  • ad hoc to paper The LLM hypothesis generator provides a reasonable prior over safety-relevant visual features.
    The generator is prompted with the same p-values used for selection in Section 3.2, so the prior is not independent of the outcome.
  • domain assumption The crash rate definition in Eq. 6 correctly adjusts for traffic exposure.
    AADT and crash counts are taken from public data; errors in these inputs propagate into all subsequent regression findings.

how reviews work

0 comments
Cite this review

Pith. "Pith review of From Street Views to Urban Science: Discovering Road Safety Factors with Multimodal Large Language Models." pith.science (2026). https://pith.science/paper/LZXQMWRU

@misc{pith2026250602242,
  author       = {Pith},
  title        = {Pith review of: From Street Views to Urban Science: Discovering Road Safety Factors with Multimodal Large Language Models},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/LZXQMWRU}},
  note         = {Machine review of arXiv:2506.02242}
}
read the original abstract

Urban and transportation research has long sought to uncover statistically meaningful relationships between key variables and societal outcomes such as road safety, to generate actionable insights that guide the planning, development, and renewal of urban and transportation systems. However, traditional workflows face several key challenges: (1) reliance on human experts to propose hypotheses, which is time-consuming and prone to confirmation bias; (2) limited interpretability, particularly in deep learning approaches; and (3) underutilization of unstructured data that can encode critical urban context. Given these limitations, we propose a Multimodal Large Language Model (MLLM)-based approach for interpretable hypothesis inference, enabling the automated generation, evaluation, and refinement of hypotheses concerning urban context and road safety outcomes. Our method leverages MLLMs to craft safety-relevant questions for street view images (SVIs), extract interpretable embeddings from their responses, and apply them in regression-based statistical models. UrbanX supports iterative hypothesis testing and refinement, guided by statistical evidence such as coefficient significance, thereby enabling rigorous scientific discovery of previously overlooked correlations between urban design and safety. Experimental evaluations on Manhattan street segments demonstrate that our approach outperforms pretrained deep learning models while offering full interpretability. Beyond road safety, UrbanX can serve as a general-purpose framework for urban scientific discovery, extracting structured insights from unstructured urban data across diverse socioeconomic and environmental outcomes. This approach enhances model trustworthiness for policy applications and establishes a scalable, statistically grounded pathway for interpretable knowledge discovery in urban and transportation studies.

Figures

Figures reproduced from arXiv: 2506.02242 by the authors.

Figure 1
Figure 1. UrbanX: an interpretable, MLLM-powered framework for hypothesis-driven urban scientific discovery. strong predictive performance, they often operate through latent feature spaces that hinder interpretability and reduce scientific transparency Lipton (2018). Third, unstructured urban data, particularly street view imagery (SVI), remains an underutilized source of contextual information. Although SVIs contain rich vis… view at source ↗
Figure 2
Figure 2. The UrbanX framework consists of three iterative modules: (1) Hypothesis Generation using LLMs, (2) Embedding Construction via MLLM-based VQA on SVIs, and (3) Hypothesis Assessment using interpretable regression analysis. 3.1. Overview Let  = {(𝑥𝑖 , 𝑦𝑖 )}𝑛 𝑖=1 denote a dataset of 𝑛 SVIs 𝑥𝑖 and their associated road-level crash rates 𝑦𝑖 ∈ ℝ. We define a hypothesis space  comprising all natural-language queries that… view at source ↗
Figure 3
Figure 3. Performance comparison between ResNet, ViT, and our interpretable embedding-based models using linear regression (LR) and LightGBM (LGBM). Lower MAE (↓) and RMSE (↓), and higher 𝑅2 (↑), indicate better performance. Our method achieves the best results across all three metrics. and coefficient of determination (𝑅2 ). These metrics are formally defined as follows: RMSE = √√√√ 1 𝑛 ∑𝑛 𝑖=1 (𝑦𝑖 − ̂𝑦𝑖 ) 2, (7) MAE = 1 𝑛 ∑𝑛… view at source ↗
Figures from the paper (6 more)
Figure 4
Figure 4. Figure 4: Spatial distribution of predicted (left) vs. actual (right) crash rates (log scale) across Manhattan road segments. UrbanX captures key spatial patterns in crash risk, including high-risk areas in lower and midtown Manhattan. segment level, the model appears to reprodu…
Figure 5
Figure 5. Figure 5: SHAP summary plot of the regression model with both traditional built environment variables and discovered hypotheses. The right panel maps the top hypothesis variables to their natural-language question meanings. (pedestrian signals present) reflect aspects of pedestr…
Figure 6
Figure 6. Figure 6: (Left) Each hypothesis is plotted by its average SHAP value and − log10(𝑝-value) from regression. Variables in the top left are highly significant and predictive. (Right) Pearson correlation matrix between hypotheses, showing low pairwise correlation and structural ind…
Figure 7
Figure 7. Figure 7: Robustness analysis. (Left) Convergence of test metrics across different LLM and MLLM configurations. High￾capacity MLLMs (78B) yield better and faster convergence. (Right) Performance as a function of the number of hypotheses used. Optimal performance is observed at 5…
Figure 8
Figure 8. Figure 8: Panoramic SVI used for VQA analysis [PITH_FULL_IMAGE:figures/full_fig_p015_8.png]
Figure 9
Figure 9. Figure 9: Final set of 50 natural-language hypotheses retained by UrbanX after the iterative process. Yihong Tang et al.: Preprint submitted to Elsevier Page 23 of 24 [PITH_FULL_IMAGE:figures/full_fig_p023_9.png]

Discussion (0). Sign in to comment.

Reference graph

Works this paper leans on

58 extracted references · 37 canonical work pages

  1. [1]

    L., Almeida, D., Altenschmidt, J., Altman, S., Anadkat, S., et al

    Achiam, J., Adler, S., Agarwal, S., Ahmad, L., Akkaya, I., Aleman, F. L., Almeida, D., Altenschmidt, J., Altman, S., Anadkat, S., et al. (2023). Gpt-4 technical report. arXiv preprint arXiv:2303.08774

    arXiv 2023

  2. [2]

    Acuto, M., Parnell, S., and Seto, K. C. (2018). Building a global urban science. Nature Sustainability , 1(1):2--4

    work page 2018

  3. [3]

    L., and Parikh, D

    Antol, S., Agrawal, A., Lu, J., Mitchell, M., Batra, D., Zitnick, C. L., and Parikh, D. (2015). Vqa: Visual question answering. In Proceedings of the IEEE international conference on computer vision , pages 2425--2433

    work page 2015

  4. [4]

    N., Wang, S., and Malikopoulos, A

    Bang, H., Dave, A., Tzortzoglou, F. N., Wang, S., and Malikopoulos, A. A. (2025). On mobility equity and the promise of emerging transportation systems. IEEE Transactions on Intelligent Transportation Systems

    work page 2025

  5. [5]

    Batty, M. (2024). The computable city: histories, technologies, stories, predictions . MIT Press

    work page 2024

  6. [6]

    X., Antonello, R

    Benara, V., Singh, C., Morris, J. X., Antonello, R. J., Stoica, I., Huth, A. G., and Gao, J. (2024). Crafting interpretable embeddings for language neuroscience by asking llms questions. Advances in neural information processing systems , 37:124137

    work page 2024

  7. [7]

    and Ito, K

    Biljecki, F. and Ito, K. (2021). Street view imagery in urban analytics and gis: A review. Landscape and Urban Planning , 215:104217

    work page 2021

  8. [8]

    Chen, Z., Wang, W., Cao, Y., Liu, Y., Gao, Z., Cui, E., Zhu, J., Ye, S., Tian, H., Liu, Z., et al. (2024). Expanding performance boundaries of open-source multimodal models with model, data, and test-time scaling. arXiv preprint arXiv:2412.05271

    arXiv 2024

Show all 58 references
  1. [9]

    Contributors, L. (2023). Lmdeploy: A toolkit for compressing, deploying, and serving llm. https://github.com/InternLM/lmdeploy

    2023

  2. [10]

    G., Gaggiotti, O

    Csill \'e ry, K., Blum, M. G., Gaggiotti, O. E., and Fran c ois, O. (2010). Approximate bayesian computation (abc) in practice. Trends in ecology & evolution , 25(7):410--418

    2010

  3. [11]

    Dosovitskiy, A., Beyer, L., Kolesnikov, A., Weissenborn, D., Zhai, X., Unterthiner, T., Dehghani, M., Minderer, M., Heigold, G., Gelly, S., et al. (2020). An image is worth 16x16 words: Transformers for image recognition at scale. arXiv preprint arXiv:2010.11929

    2020 arXiv

  4. [12]

    and Dumbaugh, E

    Ewing, R. and Dumbaugh, E. (2009). The built environment and traffic safety: a review of empirical evidence. Journal of Planning Literature , 23(4):347--367

    2009

  5. [13]

    and Handy, S

    Ewing, R. and Handy, S. (2009). Measuring the unmeasurable: Urban design qualities related to walkability. Journal of Urban design , 14(1):65--84

    2009

  6. [14]

    Gettys, C. F. and Fisher, S. D. (1979). Hypothesis plausibility and hypothesis generation. Organizational behavior and human performance , 24(1):93--110

    1979

  7. [15]

    Goodfellow, I., Bengio, Y., Courville, A., and Bengio, Y. (2016). Deep learning , volume 1. MIT press Cambridge

    2016

  8. [16]

    Gottweis, J., Weng, W.-H., Daryin, A., Tu, T., Palepu, A., Sirkovic, P., Myaskovsky, A., Weissenberger, F., Rong, K., Tanno, R., et al. (2025). Towards an ai co-scientist. arXiv preprint arXiv:2502.18864

    2025 arXiv

  9. [17]

    Guzman, L. A. and Bocarejo, J. P. (2017). Urban form and spatial urban equity in bogota, colombia. Transportation research procedia , 25:4491--4506

    2017

  10. [18]

    Hall, R. (2012). Handbook of transportation science , volume 23. Springer Science & Business Media

    2012

  11. [19]

    He, K., Zhang, X., Ren, S., and Sun, J. (2016). Deep residual learning for image recognition. In Proceedings of the IEEE conference on computer vision and pattern recognition , pages 770--778

    2016

  12. [20]

    Hou, Q., Huo, X., and Leng, J. (2020). A correlated random parameters tobit model to analyze the safety effects and temporal instability of factors affecting crash rates. Accident Analysis & Prevention , 134:105326

    2020

  13. [21]

    Hu, Y., Chen, L., and Zhao, Z. (2024). How does street environment affect pedestrian crash risks? a link-level analysis using street view image-based pedestrian exposure measurement. Accident Analysis & Prevention , 205:107682

    2024

  14. [22]

    Huang, W., Wang, J., and Cong, G. (2024). Zero-shot urban function inference with street view images through prompting a pretrained vision-language model. International Journal of Geographical Information Science , 38(7):1414--1442

    2024

  15. [23]

    Huh, M., Cheung, B., Wang, T., and Isola, P. (2024). The platonic representation hypothesis. arXiv preprint arXiv:2405.07987

    2024 arXiv

  16. [24]

    P., Perelman, A., Ramesh, A., Clark, A., Ostrow, A., Welihinda, A., Hayes, A., Radford, A., et al

    Hurst, A., Lerer, A., Goucher, A. P., Perelman, A., Ramesh, A., Clark, A., Ostrow, A., Welihinda, A., Hayes, A., Radford, A., et al. (2024). Gpt-4o system card. arXiv preprint arXiv:2410.21276

    2024 arXiv

  17. [25]

    Ignatius, M., Lim, J., Gottkehaskamp, B., Fujiwara, K., Miller, C., and Biljecki, F. (2024). Digital twin and wearables unveiling pedestrian comfort dynamics and walkability in cities. ISPRS Annals of Photogrammetry, Remote Sensing & Spatial Information Sciences , 10

    2024

  18. [26]

    Ke, G., Meng, Q., Finley, T., Wang, T., Chen, W., Ma, W., Ye, Q., and Liu, T.-Y. (2017). Lightgbm: A highly efficient gradient boosting decision tree. Advances in neural information processing systems , 30

    2017

  19. [27]

    Kuang, J., Shen, Y., Xie, J., Luo, H., Xu, Z., Li, R., Li, Y., Cheng, X., Lin, X., and Han, Y. (2025). Natural language understanding and inference with mllm in visual question answering: A survey. ACM Computing Surveys , 57(8):1--36

    2025

  20. [28]

    Lipton, Z. C. (2018). The mythos of model interpretability: In machine learning, the concept of interpretability is both important and slippery. Queue , 16(3):31--57

    2018

  21. [29]

    Liu, H., Li, C., Wu, Q., and Lee, Y. J. (2023). Visual instruction tuning. Advances in neural information processing systems , 36:34892--34916

    2023

  22. [30]

    L., Ordonez-Hurtado, R., Zayats, M., Mulligan, N., and Bettencourt-Silva, J

    Lopez, V., Hoang, L., Martinez-Galindo, M., Fern \'a ndez-D \' az, R., Sbodio, M. L., Ordonez-Hurtado, R., Zayats, M., Mulligan, N., and Bettencourt-Silva, J. (2025). Enhancing foundation models for scientific discovery via multimodal knowledge graph representations. Journal o...

    2025

  23. [31]

    Lundberg, S. M. and Lee, S.-I. (2017). A unified approach to interpreting model predictions. Advances in neural information processing systems , 30

    2017

  24. [32]

    and Qian, Z

    Ma, W. and Qian, Z. S. (2018a). Estimating multi-year 24/7 origin-destination demand using high-granular multi-source traffic data. Transportation Research Part C: Emerging Technologies , 96:96--121

    2018

  25. [33]

    and Qian, Z

    Ma, W. and Qian, Z. S. (2018b). Statistical inference of probabilistic origin-destination demand using day-to-day traffic data. Transportation Research Part C: Emerging Technologies , 88:227--256

    2018

  26. [34]

    A., Kwasigroch, A., and Majek, K

    Majchrowska, S., Miko ajczyk, A., Ferlin, M., Klawikowska, Z., Plantykow, M. A., Kwasigroch, A., and Majek, K. (2022). Deep learning-based waste detection in natural and urban environments. Waste Management , 138:274--284

    2022

  27. [35]

    C., Peck, E

    Montgomery, D. C., Peck, E. A., and Vining, G. G. (2021). Introduction to linear regression analysis . John Wiley & Sons

    2021

  28. [36]

    Moon, T. K. (1996). The expectation-maximization algorithm. IEEE Signal processing magazine , 13(6):47--60

    1996

  29. [37]

    U., Qiu, S., Saqib, M., Anwar, S., Usman, M., Akhtar, N., Barnes, N., and Mian, A

    Naveed, H., Khan, A. U., Qiu, S., Saqib, M., Anwar, S., Usman, M., Akhtar, N., Barnes, N., and Mian, A. (2023). A comprehensive overview of large language models. arXiv preprint arXiv:2307.06435

    2023 arXiv

  30. [38]

    Nie, T., Sun, J., and Ma, W. (2025). Exploring the roles of large language models in reshaping transportation systems: A survey, framework, and roadmap. arXiv preprint arXiv:2503.21411

    2025 arXiv

  31. [39]

    W., Hallacy, C., Ramesh, A., Goh, G., Agarwal, S., Sastry, G., Askell, A., Mishkin, P., Clark, J., et al

    Radford, A., Kim, J. W., Hallacy, C., Ramesh, A., Goh, G., Agarwal, S., Sastry, G., Askell, A., Mishkin, P., Clark, J., et al. (2021). Learning transferable visual models from natural language supervision. In International conference on machine learning , pages 8748--8763. PmLR

    2021

  32. [40]

    Ruder, S. (2016). An overview of gradient descent optimization algorithms. arXiv preprint arXiv:1609.04747

    2016 arXiv

  33. [41]

    Santamouris, M. (2013). Energy and climate in the urban built environment . Routledge

    2013

  34. [42]

    G., Numminen, E., Corander, J., Foll, M., and Dessimoz, C

    Sunn ker, M., Busetto, A. G., Numminen, E., Corander, J., Foll, M., and Dessimoz, C. (2013). Approximate bayesian computation. PLoS computational biology , 9(1):e1002803

    2013

  35. [43]

    Tang, Y., Kong, M., and Sun, L. (2025). Large language models for data synthesis. arXiv preprint arXiv:2505.14752

    2025 arXiv

  36. [44]

    Tang, Y., Wang, Z., Qu, A., Yan, Y., Wu, Z., Zhuang, D., Kai, J., Hou, K., Guo, X., Zhao, J., et al. (2024). Itinera: Integrating spatial optimization with large language models for open-domain urban itinerary planning. In Proceedings of the 2024 Conference on Empirical Method...

    2024

  37. [45]

    M., et al

    Thawakar, O., Dissanayake, D., More, K., Thawkar, R., Heakl, A., Ahsan, N., Li, Y., Zumri, M., Lahoud, J., Anwer, R. M., et al. (2025). Llamav-o1: Rethinking step-by-step visual reasoning in llms. arXiv preprint arXiv:2501.06186

    2025 arXiv

  38. [46]

    Wang, Z., Zhu, X., Yang, X., Luo, G., Li, H., Tian, C., Dou, W., Ge, J., Lu, L., Qiao, Y., et al. (2025). Parameter-inverted image pyramid networks for visual perception and multimodal understanding. arXiv preprint arXiv:2501.07783

    2025 arXiv

  39. [47]

    Wei, J., Tay, Y., Bommasani, R., Raffel, C., Zoph, B., Borgeaud, S., Yogatama, D., Bosma, M., Zhou, D., Metzler, D., et al. (2022a). Emergent abilities of large language models. arXiv preprint arXiv:2206.07682

    2022 arXiv

  40. [48]

    V., Zhou, D., et al

    Wei, J., Wang, X., Schuurmans, D., Bosma, M., Xia, F., Chi, E., Le, Q. V., Zhou, D., et al. (2022b). Chain-of-thought prompting elicits reasoning in large language models. Advances in neural information processing systems , 35:24824--24837

    2022

  41. [49]

    and Lee, J

    Wong, W. and Lee, J. (2005). Statistical analysis of geographic information with ArcView GIS and ArcGIS . Wiley

    2005

  42. [50]

    Wu, J., Gan, W., Chen, Z., Wan, S., and Yu, P. S. (2023). Multimodal large language models: A survey. In 2023 IEEE International Conference on Big Data (BigData) , pages 2247--2256. IEEE

    2023

  43. [51]

    Xia, Y., Qu, A., Zheng, Y., Tang, Y., Zhuang, D., Liang, Y., Wang, S., Wu, C., Sun, L., Zimmermann, R., and Zhao, J. (2025). Reimagining urban science: Scaling causal inference with large language models. arXiv preprint arXiv:2504.12345

    2025 arXiv

  44. [52]

    Xue, H., Guo, P., Li, Y., and Ma, J. (2024). Integrating visual factors in crash rate analysis at intersections: An automl and shap approach towards cycling safety. Accident Analysis & Prevention , 200:107544

    2024

  45. [53]

    Yang, Z., Li, L., Lin, K., Wang, J., Lin, C.-C., Liu, Z., and Wang, L. (2023). The dawn of lmms: Preliminary explorations with gpt-4v (ision). arXiv preprint arXiv:2309.17421 , 9(1):1

    2023 arXiv

  46. [54]

    Yu, X., Ma, J., Tang, Y., Yang, T., and Jiang, F. (2024). Can we trust our eyes? interpreting the misperception of road safety from street view images and deep learning. Accident Analysis & Prevention , 197:107455

    2024

  47. [55]

    Yue, Y., Chen, Z., Lu, R., Zhao, A., Wang, Z., Song, S., and Huang, G. (2025). Does reinforcement learning really incentivize reasoning capacity in llms beyond the base model? arXiv preprint arXiv:2504.13837

    2025 arXiv

  48. [56]

    Zeng, Q., Wen, H., Huang, H., Pei, X., and Wong, S. (2017). A multivariate random-parameters tobit model for analyzing highway crash rates by injury severity. Accident Analysis & Prevention , 99:184--191

    2017

  49. [57]

    Zhang, D., Yu, Y., Dong, J., Li, C., Su, D., Chu, C., and Yu, D. (2024). Mm-llms: Recent advances in multimodal large language models. arXiv preprint arXiv:2401.13601

    2024 arXiv

  50. [58]

    write newline

    " write newline "" before.all 'output.state := FUNCTION n.dashify 't := "" t empty not t #1 #1 substring "-" = t #1 #2 substring "--" = not "--" * t #2 global.max substring 't := t #1 #1 substring "-" = "-" * t #2 global.max substring 't := while if t #1 #1 substring * t #2 gl...

Pith tools

Reviewed August 7, 2026 · model on record in the stance chip above.