Pith. sign in

REVIEW 3 major objections 6 minor 47 references

SCENT: Robust Spatiotemporal Learning for Continuous Scientific Data via Scalable Conditioned Neural Fields

T0 review · 3 major / 6 minor · reviewed 2026-08-16 · deepseek-v4-flash

Pith's one-line read SCENT is a single transformer-based neural field that reconstructs, interpolates, and forecasts sparse scientific data from one trained model, and the paper reports it outranks every baseline on all three Navier-Stokes benchmarks.

desk verdict Useful unified neural-field method for irregular spatiotemporal data, but the headline long-horizon SOTA claim rests on an unablated test-time warp-unrolling advantage. read the letter →

arxiv 2504.12262 v1 pith:PRWIZLIK submitted 2025-04-16 cs.LG cs.AI

classification cs.LGcs.AI
keywords conditionedneuralfieldsspatiotemporalforecastingimplicitrepresentationslearnablequeriessparseattentionwarp-unrollingNavier-StokesbenchmarksPM2.5airquality
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

The paper sets out to establish that one architecture—a transformer-based conditioned neural field named SCENT—can learn a continuous spatiotemporal field from sparse, noisy, moving sensor readings and then reconstruct, spatially interpolate, and forecast it from a single trained model. The point would matter because scientific data often arrives exactly like that: malfunctioning air-quality monitors, satellite pixels, and simulation output on irregular meshes, which today need separate pipelines for cleaning, filling in, and predicting. The paper further claims that SCENT outperforms every baseline on all three Navier-Stokes benchmarks and on seven of the eight simulated and real sensor regimes it tests, and that its error keeps falling as model and dataset size grow, where a Fourier neural operator's curve flattens. On a sympathetic reading, SCENT is a single tool for continuous spatiotemporal fields that adapts to whatever sensors exist at test time.

What carries the argument

The load-bearing object is the encoder-processor-decoder conditioned neural field built on $M$ learnable query tokens (the inducing-point pattern from Perceiver-style cross-attention architectures). The encoder's Context Embedding Network applies sparse self-attention in which each of the $N$ input tokens attends to a random subset of $S \ll N$ tokens before cross-attending against the $M$ queries, so cost is linear in input size and the model accepts any $N$; the Time Warp Processor moves the queries through time by a continuous $\Delta t$; the decoder's Calibration Network maps the queries back to arbitrary coordinates with Fourier features and sparse self-attention. Warp-Unrolling Forecasting is the inference mechanism that carries the long-horizon results: instead of stepping one time unit at a time, the model advances directly to the training horizon $t_h$ and uses that state as the reference for the remaining steps, so at any state only a minimal number of prediction steps remain.

What would settle it

Re-run the S1–S5 and AirDelhi comparisons with every baseline given the same per-dataset validation-based hyperparameter search and matched compute as SCENT; if FNO or AROMA then matches or beats SCENT on S2 or AD-T, the universal-outperformance claim collapses, and an ablation that swaps warp-unrolling for one-step unrolling at identical training would settle whether WUF, rather than the architecture itself, produces the NS-3 win.

Watch

Extended reading notes

Core claim

SCENT parameterizes the target field as a function of space-time coordinates conditioned on input values: Fourier features encode the coordinates $(x,t)$, input samples are linearly projected, and a cross-attention encoder compresses any number $N$ of input samples into $M$ learnable query tokens; a Time Warp Processor shifts those tokens from input time $t_i$ to target time $t_o$ by a continuous step $\Delta t \in [0, t_h]$, and a time-conditioned decoder evaluates the field at arbitrary output locations through Fourier features and sparse self-attention. The same forward pass performs reconstruction ($\Delta t = 0$), interpolation at novel locations or times, and forecasting ($t_o > t_i$), so no latent optimization or meta-learning is needed per dataset. Two devices carry the empirical results: the Context Embedding and Calibration Networks add sparse self-attention at input and output, and warp-unrolling forecasting jumps to the horizon $t_h$ in one step instead of unrolling every tick, which the paper credits for the large NS-3 gain. The reported scores beat FNO, OFormer, DINO, CORAL, and AROMA on all three Navier-Stokes benchmarks and on the simulated S1–S5 variants, and beat all baselines on the two fine-grained AirDelhi datasets; on the coarsest AirDelhi variant SCENT places second after AROMA, the paper's single acknowledged exception.

Load-bearing premise

The load-bearing premise is that the baselines were fairly adapted and tuned for the irregular regimes—the paper never specifies how FNO, OFormer, CORAL, or AROMA were configured on each dataset, and the S2 margin is only $2.08 \times 10^{-1}$ versus $2.10 \times 10^{-1}$, so a weaker baseline setup could reverse the reported ordering.

Editorial extensions

If this is right

  • One trained SCENT model replaces three separate tools—reconstruction, interpolation, and forecasting—so sensor cleaning and prediction no longer have to be staged pipelines.
  • Long-horizon forecasts on slow dynamics improve sharply: on NS-3 the paper reports MSE $7.78 \times 10^{-5}$ versus $1.32 \times 10^{-4}$ for the best prior model, with the gap attributed to warp-unrolling's reduction of error accumulation.
  • The model accepts a variable number of input and output locations in one forward pass, which the paper states no baseline handles naturally; this covers missing sensors, moving sensors, and arbitrary output resolution.
  • Larger models and datasets keep paying off: the scalability study shows SCENT's error falling along a linear trend while FNO's converges, and 100k training trajectories beat 30k.
  • Because $\Delta t$ is sampled uniformly in $[0, t_h]$ during training, predictions are available at any continuous time offset inside the horizon, not just at integer steps.

Reading between the lines

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

  • A stress test the paper leaves implicit: train on one sensor-count regime and test with far fewer or far more sensors than were seen in training, to see how far the 'variable $N_i$' claim extends beyond in-distribution interpolation.
  • Sparse attention makes each token's receptive field a random sample of $S$ neighbors; ablating $S$ from a handful of neighbors up to full attention would reveal whether the continuity gains come from global context or from local field smoothness.
  • The same conditioning mechanism could ingest asynchronous streams in which every measurement carries its own timestamp; the time-warp training already permits non-integer intervals, so event-driven sensor feeds are a direct testbed.
  • The complexity analysis implies SCENT's inference cost scales with $W \approx T/t_h$ rather than with total steps $T$, so its advantage over FNO and AROMA should widen on very long horizons; a wall-clock comparison at matched error would make that concrete.
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

3 major / 6 minor

Summary. The paper introduces SCENT, a conditioned neural field with a transformer-based encoder-processor-decoder, learnable latent queries, sparse self-attention, time-targeted encoding, and a warp-unrolling forecasting (WUF) inference strategy. The authors claim a single-stage model that jointly performs reconstruction, interpolation, and forecasting on sparse, noisy, moving-sensor data, outperforming FNO, OFormer, CORAL, AROMA, DINO, and GNOT on simulated Navier-Stokes variants, three Navier-Stokes benchmarks, and three AirDelhi variants, while scaling better than FNO. The paper provides pseudo-code, hyperparameter tables, ablations of the architectural components, and a complexity analysis.

Significance. SCENT addresses a relevant problem: learning continuous spatiotemporal fields from irregularly sampled scientific data. The paper contributes a broad empirical study with five simulated corruption types, three Navier-Stokes benchmarks, three AirDelhi variants, architectural ablations, and a scalability analysis, and it provides a pseudo-algorithm and detailed hyperparameter tables that are useful for reproducibility. However, the headline performance claims are weakened by an uncontrolled asymmetry in the inference protocol (WUF is applied only to SCENT) and by the absence of error bars on differences that are often small. If the WUF confound is resolved and the comparisons are shown to be statistically robust, the contribution would be solid; under the current evidence, the central 'outperforms all baselines' assertion is not established.

major comments (3)
  1. [Section 4.5 and Section 3.3] The Table 2 claim that 'SCENT outperforms all baseline models across all datasets' is confounded by an unablated change in the evaluation protocol. All models are trained with next-state supervision, but SCENT is evaluated using WUF, which advances directly by up to th=5 steps, whereas the baselines must unroll one step at a time (Section 3.3). The paper itself attributes the advantage to WUF: 'This advantage is particularly evident... which we attribute to WUF, fundamentally enabled by the time-continuity learned by the model.' To establish that the learned representation is better, the authors must either evaluate SCENT using standard one-step unrolling, or train the baselines with multi-step supervision up to th and give them the same warp-unrolling inference. Without such a control, Table 2 reflects an inference-protocol advantage rather than an isolated model-quality advantage.
  2. [Tables 1 and 2] No error bars, standard deviations, or repeated-seed results are reported anywhere in the paper. Several margins that support the 'consistently outperforms' claim are very small: S2 (2.08e-1 vs 2.10e-1), NS-4 (1.03e-1 vs 1.05e-1), and NS-5 (1.17e-1 vs 1.24e-1). Without at least 3-5 seeds and a statement of variance or significance, these differences could be run-to-run noise. Please report mean plus/minus standard deviation and, ideally, a paired significance test for the main comparisons.
  3. [Section 4.1.3 and Section 4.2 (baselines)] The baseline adaptation for irregular and moving sensors is underspecified. Section 4.2 describes only how FNO is modified (zero-padding plus a mask on the loss), and Section 4.1.3 states that SCENT is the only model that can 'naturally handle a variable Ni and No,' which raises the question of how OFormer, CORAL, and AROMA are configured for datasets S5, AD-B, AD-T, and AD-F. Please document for each baseline the input featurization on irregular or moving coordinates, any architecture modifications, hyperparameter search budgets, and the selection criteria (e.g., a chosen validation set). This is needed to rule out that the reported ranking comes from suboptimally adapted baselines.
minor comments (6)
  1. [Appendix L, Table 6] For a 20-step horizon with th=5, WUF requires about 4 forward passes, but Table 6 lists W=7; clarify how W is counted.
  2. [Section 3.2] The phrase 'linear projection layer with parameters frozen' should clarify whether the projection is pre-trained and kept fixed, and why that choice is made.
  3. [Figure 5 caption] 'delta' appears in the figure but is not defined in the caption; please define what delta represents.
  4. [Table 3] The 'CONTRAST' column should state explicitly that the percentages are relative degradations in Rel-MSE with respect to the full model.
  5. [Appendices C and D] The data-statistics row labeled 'N POINTS - INPUTS (M)' appears to be a typo; if M denotes the output points No, rename it for clarity.
  6. [Section 3.3] The heading 'Temporal Warp Processor' and the later phrase 'Time Warp Processor' are used inconsistently; pick one term and use it throughout.

Circularity Check

0 steps flagged · score 0.0 of 10

No load-bearing circularity: SCENT's benchmark claims rest on external data and held-out metrics; the WUF asymmetry is an evaluation concern, not a circular derivation.

full rationale

The paper's headline results are computed on held-out test trajectories of external Navier-Stokes benchmarks (NS-3/4/5), the public AirDelhi real-world dataset, and newly generated but physically simulated variants (S1–S5); evaluation metrics are defined in Eq. 1 and the comparisons are against external baselines. No parameter is fitted to the reported test values, and no prediction is obtained by algebraic rearrangement of the model's own assumptions. The one self-citation (Lee & Oh 2024 in Section 3.1) supplies architectural context alongside independent references such as Jaegle et al. 2022; it is not used to justify any empirical claim and is therefore not load-bearing. The WUF protocol in Section 4.5 — SCENT is evaluated with multi-step warp-unrolling while next-step baselines unroll one step at a time, and no WUF ablation is reported — is a legitimate evaluation-protocol confound and a correctness/fairness risk, but it is not an equation-level reduction or a fitted parameter renamed as a prediction. The ablations in Table 3 are empirical and internal, not definitional. Overall, the model is a learned map from input observations to outputs, and its 'predictions' are measured on data not used to fit the model, so the derivation chain is not circular.

Assumptions & free parameters 5 free parameters · 4 assumptions · 0 invented entities

The central contribution is empirical. The main entries are hand-chosen hyperparameters that tune the model to each dataset, plus three domain assumptions about simulator fidelity, WUF time-generalization, and baseline fairness. No new physical or ontological entities are introduced.

free parameters (5)
  • time_horizon th = 3 (S1-S5), 5 (NS-3/4/5)
    Sets the maximum delta t used in training and the jump size in warp-unrolling forecasting; chosen per dataset and directly shapes WUF behavior.
  • number of learnable queries M = 64 (NS-3), 256 (NS-4/5), 128 (S1-S5), 128-256 (scalability runs)
    Controls the size of the compressed latent representation; chosen per dataset and model scale, affects accuracy.
  • latent_dim = 128 (most), up to 1024 (scalability)
    Model width; tuned per experiment, directly affects capacity and accuracy.
  • fourier_frequency_bands = 6 (NS-3), 12 (NS-4/5, S1-S5)
    Number of Fourier feature bands for coordinate encoding; affects high-frequency detail and is chosen per dataset.
  • sparse_attention_group_size = 1 or 8 (NS), 2 or 8 (S1-S5)
    Number of tokens each token attends to in sparse self-attention; balances cost and context, tuned per dataset.
assumptions (4)
  • domain assumption The Navier-Stokes simulator (vorticity transport equation with GRF initialization and forcing f = -4 cos(4x2)) produces ground-truth fields that faithfully represent the scientific regimes of interest.
    All S1-S5 and NS benchmarks are evaluated against these simulations, so simulator fidelity is assumed. Appendix E.
  • domain assumption A model trained with next-state supervision and random delta t in [0, th] can be applied at inference with WUF jumps of size th without distribution shift or error blow-up.
    WUF (Section 3.3, Appendix B) repeatedly calls the model with increment th; the paper does not report error as a function of delta t, so this generalization across the time horizon is assumed.
  • domain assumption Baseline models (FNO, OFormer, CORAL, AROMA) were implemented and tuned correctly and fairly for variable-input, moving-sensor settings.
    The SOTA claim is comparative; the paper does not specify how each baseline was adapted for variable Ni/No and moving sensors (Sections 4.1.3 and 4.2).
  • standard math The cross-attention and sparse-attention computations follow the standard definitions from the cited Perceiver IO and Transformer literature, with no unstated normalization or masking beyond what is described.
    The architecture is defined relative to prior work (Section 3.1), so correctness relies on those definitions.

how reviews work

0 comments
Cite this review

Pith. "Pith review of SCENT: Robust Spatiotemporal Learning for Continuous Scientific Data via Scalable Conditioned Neural Fields." pith.science (2026). https://pith.science/paper/PRWIZLIK

@misc{pith2026250412262,
  author       = {Pith},
  title        = {Pith review of: SCENT: Robust Spatiotemporal Learning for Continuous Scientific Data via Scalable Conditioned Neural Fields},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/PRWIZLIK}},
  note         = {Machine review of arXiv:2504.12262}
}
read the original abstract

Spatiotemporal learning is challenging due to the intricate interplay between spatial and temporal dependencies, the high dimensionality of the data, and scalability constraints. These challenges are further amplified in scientific domains, where data is often irregularly distributed (e.g., missing values from sensor failures) and high-volume (e.g., high-fidelity simulations), posing additional computational and modeling difficulties. In this paper, we present SCENT, a novel framework for scalable and continuity-informed spatiotemporal representation learning. SCENT unifies interpolation, reconstruction, and forecasting within a single architecture. Built on a transformer-based encoder-processor-decoder backbone, SCENT introduces learnable queries to enhance generalization and a query-wise cross-attention mechanism to effectively capture multi-scale dependencies. To ensure scalability in both data size and model complexity, we incorporate a sparse attention mechanism, enabling flexible output representations and efficient evaluation at arbitrary resolutions. We validate SCENT through extensive simulations and real-world experiments, demonstrating state-of-the-art performance across multiple challenging tasks while achieving superior scalability.

Figures

Figures reproduced from arXiv: 2504.12262 by the authors.

Figure 1
Figure 1. Challenging data scenarios motivating this study. (a) Learning continuous ground truth signal (GT) given noisy, mal￾functioning, sparse, or moving sensors is a daunting challenge. However, these challenges are common in scientific data, including AirDelhi (Chauhan et al., 2024) data which measure the particulate matter levels (PM2.5) from moving vehicles. with larger models and datasets. • We conduct extensive exper… view at source ↗
Figure 2
Figure 2. SCENT overview. (a) Detailed architecture of SCENT is illustrated. Our unique contributions are drawn with red boxes. Specifically, we introduce time coordinates to both encoder and decoder for learning continuous time representations. Also, we introduce Context Embedding Network and Calibration Network for improved spatial encoding and decoding, respectively. nenc, nproc, and ndec denote the number of layers in enc… view at source ↗
Figure 3
Figure 3. Warp-unrolling forecasting (WUF). (a) Conventional forecasting includes single-step unrolling which accumulates error. A dimming blue color is used to represent the increasing error. (b) WUF helps mitigate error accumulations caused by extensive unrolling steps. {xj} Ni j=1 can be structured on a grid or an irregular mesh. Using a cross-attention mechanism, the encoder transforms U ti into a fixed-size set of tokens… view at source ↗
Figures from the paper (3 more)
Figure 4
Figure 4. Figure 4: Scalability evaluations. (a) Texts next to each circle are the number of model parameters, and circle size is also proportional to it. Red dotted lines are scalability trends derived with exponential functions for comparisons. (b) Colors indicate training runs with ide…
Figure 5
Figure 5. Figure 5: Qualitative comparisons for forecasting. (a) Using partial input from the S5 dataset, each pretrained model is assessed on the full mesh (GT), performing both joint forecasting and spatial interpolation. (b) Models forecast PM2.5 on the spatiotemporal locations where m…
Figure 6
Figure 6. Figure 6: Joint reconstruction, interpolation, and forecasting. Given dataset S5 inputs shown on top, neural fields are tested for reconstruction/interpolation (ti) and forecasting at continuous time (t¯io, to). For comparison purposes, interpolation results from nearest-neighbo…

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

47 extracted references · 36 canonical work pages

  1. [1]

    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 global.max substring 't := if while FUNCTION format.date year duplicate empty "emp...

  2. [2]

    The atlas experiment at the cern large hadron collider

    Aad, G., Abbott, B., Abdallah, J., et al. The atlas experiment at the cern large hadron collider. Journal of Instrumentation, 3: 0 S08003, 2019

  3. [3]

    Multiscale mobility networks and the spatial spreading of infectious diseases

    Balcan, D., Colizza, V., et al. Multiscale mobility networks and the spatial spreading of infectious diseases. Proceedings of the National Academy of Sciences, 106 0 (51): 0 21484--21489, 2009

  4. [4]

    R., and Kim, H

    Bauer, M., Dupont, E., Brock, A., Rosenbaum, D., Schwarz, J. R., and Kim, H. Spatial functa: Scaling functa to imagenet classification and generation. arXiv preprint arXiv:2302.03130, 2023

  5. [5]

    Is space-time attention all you need for video understanding? In International Conference on Machine Learning (ICML), 2021

    Bertasius, G., Wang, H., and Torresani, L. Is space-time attention all you need for video understanding? In International Conference on Machine Learning (ICML), 2021

  6. [6]

    B., Ranu, S., Sen, R., and Batra, N

    Chauhan, S., Patel, Z. B., Ranu, S., Sen, R., and Batra, N. Airdelhi: fine-grained spatio-temporal particulate matter dataset from delhi for ml based modeling. Advances in Neural Information Processing Systems, 36, 2024

  7. [7]

    and Wang, X

    Chen, Y. and Wang, X. Transformers as meta-learners for implicit neural representations. In European Conference on Computer Vision, pp.\ 170--187. Springer, 2022

  8. [8]

    Transinr: Encoding and decoding images as implicit neural fields using data augmentation

    Chen, Z., Wang, S., Liu, C., Yang, T., Dai, B., and Lin, D. Transinr: Encoding and decoding images as implicit neural fields using data augmentation. Advances in Neural Information Processing Systems (NeurIPS), 2021

Show all 47 references
  1. [9]

    An image is worth 16x16 words: Transformers for image recognition at scale

    Dosovitskiy, A., Beyer, L., Kolesnikov, A., Weissenborn, D., Zhai, X., Unterthiner, T., Dehghani, M., Minderer, M., Heigold, G., Gelly, S., et al. An image is worth 16x16 words: Transformers for image recognition at scale. In International Conference on Learning Representation...

  2. [10]

    Dupont, E., Kim, H., Eslami, S. M. A., Rezende, D. J., and Rosenbaum, D. From data to functa: Your data point is a function and you can treat it like one. In International Conference on Machine Learning, pp.\ 5694--5725. PMLR, 2022

  3. [11]

    S., Tonolini, F., and Murray-Smith, R

    Gabbard, H., Messenger, C., Heng, I. S., Tonolini, F., and Murray-Smith, R. Bayesian parameter estimation using conditional variational autoencoders for gravitational-wave astronomy. Nature Physics, 18 0 (1): 0 112--117, 2022

  4. [12]

    and Schmidhuber, J

    Hochreiter, S. and Schmidhuber, J. Long short-term memory. Neural Computation, 9 0 (8): 0 1735--1780, 1997

  5. [13]

    J., Botvinick, M., Zisserman, A., Vinyals, O., and Carreira, J

    Jaegle, A., Borgeaud, S., Alayrac, J.-B., Doersch, C., Ionescu, C., Ding, D., Koppula, S., Zoran, D., Brock, A., Shelhamer, E., Henaff, O. J., Botvinick, M., Zisserman, A., Vinyals, O., and Carreira, J. Perceiver IO : A general architecture for structured inputs & outputs. In ...

  6. [14]

    Highly accurate protein structure prediction with alphafold

    Jumper, J., Evans, R., Pritzel, A., et al. Highly accurate protein structure prediction with alphafold. Nature, 596: 0 583--589, 2021

  7. [15]

    Kaltenborn, J., Lange, C. E. E., Ramesh, V., Brouillard, P., Gurwicz, Y., Nagda, C., Runge, J., Nowack, P., and Rolnick, D. Climateset: A large-scale climate model dataset for machine learning. In Thirty-seventh Conference on Neural Information Processing Systems Datasets and ...

  8. [16]

    Digital typhoon: Long-term satellite image dataset for the spatio-temporal modeling of tropical cyclones

    Kitamoto, A., Hwang, J., Vuillod, B., Gautier, L., Tian, Y., and Clanuwat, T. Digital typhoon: Long-term satellite image dataset for the spatio-temporal modeling of tropical cyclones. In Thirty-seventh Conference on Neural Information Processing Systems Datasets and Benchmarks...

  9. [17]

    Lee, D., Kim, C., Cho, M., and HAN, W. S. Locality-aware generalizable implicit neural representation. Advances in Neural Information Processing Systems, 36: 0 48363--48381, 2023

  10. [18]

    and Oh, T

    Lee, S. and Oh, T. Inducing point operator transformer: A flexible and scalable architecture for solving pdes. In Proceedings of the AAAI Conference on Artificial Intelligence, volume 38, pp.\ 153--161, 2024

  11. [19]

    Diffusion convolutional recurrent neural network: Data-driven traffic forecasting

    Li, Y., Yu, R., Shahabi, C., and Liu, Y. Diffusion convolutional recurrent neural network: Data-driven traffic forecasting. In International Conference on Learning Representations (ICLR), 2018

  12. [20]

    Fourier neural operator for parametric partial differential equations

    Li, Z., Kovachki, N., Azizzadenesheli, K., Liu, B., Bhattacharya, K., Stuart, A., and Anandkumar, A. Fourier neural operator for parametric partial differential equations. arXiv preprint arXiv:2010.08895, 2020

  13. [21]

    Neural fields for pde solving

    Li, Z., Kovachki, N., Azizzadenesheli, K., Bhattacharya, K., Anandkumar, A., and Stuart, A. Neural fields for pde solving. Neural Networks, 2022

  14. [22]

    Li, Z., Meidani, K., and Farimani, A. B. Transformer for partial differential equations operator learning. Transactions on Machine Learning Research, 2023. ISSN 2835-8856. URL https://openreview.net/forum?id=EPPqt3uERT

  15. [23]

    E., Setio, A

    Litjens, G., Kooi, T., Bejnordi, B. E., Setio, A. A. A., Ciompi, F., Ghafoorian, M., Van Der Laak, J. A., Van Ginneken, B., and Sánchez, C. I. A survey on deep learning in medical image analysis. Medical Image Analysis, 42: 0 60--88, 2017

  16. [24]

    A., Wheeler, J., Beltr \'a n-Deb \'o n, R., Joven, J., Sales-Pardo, M., and Guimer \`a , R

    Massucci, F. A., Wheeler, J., Beltr \'a n-Deb \'o n, R., Joven, J., Sales-Pardo, M., and Guimer \`a , R. Inferring propagation paths for sparsely observed perturbations on complex networks. Science advances, 2 0 (10): 0 e1501638, 2016

  17. [25]

    and Marusic, I

    Meneveau, C. and Marusic, I. Turbulence in the atmosphere. Annual Review of Fluid Mechanics, 43: 0 219--245, 2011

  18. [26]

    P., Tancik, M., Barron, J

    Mildenhall, B., Srinivasan, P. P., Tancik, M., Barron, J. T., Ramamoorthi, R., and Ng, R. Nerf: Representing scenes as neural radiance fields for view synthesis. In European Conference on Computer Vision (ECCV), pp.\ 405--421. Springer, 2020

  19. [27]

    B., Dong, J., Tucker, C

    Myneni, R. B., Dong, J., Tucker, C. J., Kaufmann, R. K., Kauppi, P. E., Liski, J., Zhou, L., Alexeyev, V., and Hughes, M. A large carbon sink in the woody biomass of northern forests. Proceedings of the National Academy of Sciences, 98 0 (26): 0 14784--14789, 2001

  20. [28]

    Let's share commrad: Effect of radar interference on an uncoded data communication system

    Nartasilpa, N., Tuninetti, D., Devroye, N., and Erricolo, D. Let's share commrad: Effect of radar interference on an uncoded data communication system. In 2016 IEEE Radar Conference (RadarConf), pp.\ 1--5. IEEE, 2016

  21. [29]

    Metaflux: Meta-learning global carbon fluxes from sparse spatiotemporal observations

    Nathaniel, J., Liu, J., and Gentine, P. Metaflux: Meta-learning global carbon fluxes from sparse spatiotemporal observations. Scientific Data, 10 0 (1): 0 440, 2023

  22. [30]

    Spatiotemporal implicit neural representation as a generalized traffic data learner

    Nie, T., Qin, G., Ma, W., and Sun, J. Spatiotemporal implicit neural representation as a generalized traffic data learner. Transportation Research Part C: Emerging Technologies, 153: 0 104890, 2024

  23. [31]

    N., Carpov, D., Chapados, N., and Bengio, Y

    Oreshkin, B. N., Carpov, D., Chapados, N., and Bengio, Y. N-beats: Neural basis expansion analysis for interpretable time series forecasting. In International Conference on Learning Representations, 2020. URL https://openreview.net/forum?id=r1ecqn4YwB

  24. [32]

    Improved particle-flow event reconstruction with scalable neural networks for current and future particle detectors

    Pata, J., Wulff, E., Mokhtar, F., Southwick, D., Zhang, M., Girone, M., and Duarte, J. Improved particle-flow event reconstruction with scalable neural networks for current and future particle detectors. Communications Physics, 7 0 (1): 0 124, 2024

  25. [33]

    B., Purohit, P., Patel, H

    Patel, Z. B., Purohit, P., Patel, H. M., Sahni, S., and Batra, N. Accurate and scalable gaussian processes for fine-grained air quality inference. In Proceedings of the AAAI Conference on Artificial Intelligence, volume 36, pp.\ 12080--12088, Jun. 2022. doi:10.1609/aaai.v36i11...

  26. [34]

    Peters, A. J. and Janyst, L. Exabyte scale storage at cern. In Journal of Physics: Conference Series, volume 331, pp.\ 052015. IOP Publishing, 2011

  27. [35]

    Deep learning and process understanding for data-driven earth system science

    Reichstein, M., Camps-Valls, G., Stevens, B., Jung, M., Denzler, J., Carvalhais, N., and Prabhat, F. Deep learning and process understanding for data-driven earth system science. Nature, 566 0 (7743): 0 195--204, 2019

  28. [36]

    Koupa \

    Serrano, L., Le Boudec, L., Kassa \" Koupa \" , A., Wang, T. X., Yin, Y., Vittaut, J.-N., and Gallinari, P. Operator learning with neural fields: Tackling pdes on general geometries. Advances in Neural Information Processing Systems, 36: 0 70581--70611, 2023

  29. [37]

    X., Naour, E

    Serrano, L., Wang, T. X., Naour, E. L., Vittaut, J.-N., and Gallinari, P. AROMA : Preserving spatial structure for latent PDE modeling with local neural fields. In The Thirty-eighth Annual Conference on Neural Information Processing Systems, 2024. URL https://openreview.net/fo...

  30. [38]

    B., and Wetzstein, G

    Sitzmann, V., Lindell, D. B., and Wetzstein, G. Metasdf: Meta-learning signed distance functions. Advances in Neural Information Processing Systems (NeurIPS), 33: 0 10136--10147, 2020 a

  31. [39]

    N., Bergman, A

    Sitzmann, V., Martel, J. N., Bergman, A. W., Lindell, D. B., and Wetzstein, G. Implicit neural representations with periodic activation functions. Advances in Neural Information Processing Systems (NeurIPS), 33: 0 7462--7473, 2020 b

  32. [40]

    Fourier features let networks learn high frequency functions in low dimensional domains

    Tancik, M., Srinivasan, P., Mildenhall, B., Fridovich-Keil, S., Raghavan, N., Singhal, U., Ramamoorthi, R., Barron, J., and Ng, R. Fourier features let networks learn high frequency functions in low dimensional domains. Advances in neural information processing systems, 33: 0 ...

  33. [41]

    C., Ugurbil, K., et al

    Van Essen, D. C., Ugurbil, K., et al. Exploring the human connectome. Neuron, 70 0 (5): 0 713--720, 2010

  34. [42]

    H., Sankaran, S., Wang, H., Pappas, G

    Wang, S., Seidman, J. H., Sankaran, S., Wang, H., Pappas, G. J., and Perdikaris, P. Bridging operator learning and conditioned neural fields: A unifying perspective. CoRR, abs/2405.13998, 2024. URL https://doi.org/10.48550/arXiv.2405.13998

  35. [43]

    Deep learning for video recognition: A survey

    Wang, X., Girshick, R., Gupta, A., and He, K. Deep learning for video recognition: A survey. IEEE Transactions on Pattern Analysis and Machine Intelligence, 2020

  36. [44]

    Woodcock, C. E. et al. Globcover: Esa’s global land cover map. Remote Sensing of Environment, 112: 0 1951--1962, 2008

  37. [45]

    Continuous PDE dynamics forecasting with implicit neural representations

    Yin, Y., Kirchmeyer, M., Franceschi, J.-Y., Rakotomamonjy, A., and patrick gallinari. Continuous PDE dynamics forecasting with implicit neural representations. In The Eleventh International Conference on Learning Representations, 2023. URL https://openreview.net/forum?id=B73niNjbPs

  38. [46]

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

    Yu, B., Yin, H., and Zhu, Z. Spatio-temporal graph convolutional networks: A deep learning framework for traffic forecasting. In International Joint Conference on Artificial Intelligence (IJCAI), 2017

  39. [47]

    A., Gupta, R., et al

    Yu, S., Hannah, W., Peng, L., Lin, J., Bhouri, M. A., Gupta, R., et al. Climsim: A large multi-scale dataset for hybrid physics- ML climate emulation. In Thirty-seventh Conference on Neural Information Processing Systems Datasets and Benchmarks Track, 2023. URL https://openrev...

Pith tools

Reviewed August 16, 2026 · model on record in the stance chip above.