Pith. sign in

REVIEW 3 major objections 5 minor 48 references

Locality-aware Surrogates for Gradient-based Black-box Optimization

T0 review · 3 major / 5 minor · reviewed 2026-08-09 · deepseek-v4-flash

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

Pith's one-line read Training a surrogate on nearest-neighbor value differences aligns its Jacobian with the black-box function, giving better gradients for optimization.

desk verdict A clean difference-matching surrogate loss with consistent empirical gains, but the theory overreaches: the Jacobian-alignment guarantee is proven only for the full pair loss, not the k-NN loss actually optimized. read the letter →

arxiv 2501.19161 v1 pith:GL36STDV submitted 2025-01-31 cs.LG

Ali Momeni , Stefan Uhlich , Arun Venkitaraman , Chia-Yu Hsieh , Andrea Bonetti , Ryoga Matsuo , Eisaku Ohbuchi , Lorenzo Servadei This is my paper · ORCID
classification cs.LG
keywords black-boxoptimizationsurrogatemodelgradientestimationGradPIElosslocality-awaretrainingactivelearningtheoremnearest-neighborpairs
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 tries to establish that a simple change to surrogate training produces substantially better gradient estimates for black-box optimization. Instead of matching predicted outputs pointwise, the proposed GradPIE loss compares, for each input and its nearest neighbors, the difference in predicted outputs with the difference in observed black-box outputs. The paper argues, via the gradient theorem, that minimizing this pairwise loss is equivalent to minimizing the gap between the Jacobians of the surrogate and the black-box function. Since optimization uses surrogate gradients in the backward pass, better Jacobian alignment translates into more efficient search over non-differentiable simulators. Experiments on coupled nonlinear oscillators, an operational amplifier circuit, and a 3600-dimensional optical wave-shaping system support this: the locality-aware surrogate outperforms the same architecture trained with MAE or MSE, and it reaches the baseline's quality with roughly 40-70% fewer black-box queries.

What carries the argument

The load-bearing object is the GradPIE loss itself: $\mathcal{L}_{\mathrm{GradPIE}}(\theta) = \mathbb{E}_{x \sim \mathcal{D}} \frac{1}{K} \sum_{k=1}^{K} \left| (F(x) - F(x'_k)) - (\hat{F}(x;\theta) - \hat{F}(x'_k;\theta)) \right|$, where $x'_k$ is the $k$-th nearest neighbor of $x$. Unlike a pointwise MAE or MSE loss, this loss compares value differences, so an additive constant in either function cancels; computed locally, it forces the surrogate to reproduce the same relative changes in output that the black-box shows. The gradient theorem then converts this pairwise consistency into an integral of Jacobian differences, which is the mechanism behind the claim.

What would settle it

On a smooth test function with exactly computable Jacobians, train two surrogates with the GradPIE loss using small-radius and large-radius neighbor pairs; if the large-radius model shows decreasing loss while the Jacobian error $\|J[\hat{F}] - J[F]\|$ stays flat or grows, the stated equivalence fails outside the small-radius regime. A cheaper check is to measure actual nearest-neighbor distances in the 3600-dimensional OWMS dataset: if typical distances exceed the linearization scale, the theoretical guarantee does not apply.

Watch

Extended reading notes

Core claim

The central claim, in the paper's own terms, is that minimizing the GradPIE loss in Eq. (5), computed over k-nearest-neighbor pairs in Eq. (10), corresponds to minimizing the difference between the Jacobians of the surrogate and the black-box function. Starting from the gradient theorem, the loss is rewritten as an integral of the Jacobian difference along the curve between paired inputs; under a locality assumption $\|x - x'\| \le \epsilon$, this integral is proportional to the row-wise norm of $J[\hat{F}](x;\theta) - J[F](x)$. Hence GradPIE-trained surrogates provide gradient estimates that better match the true gradients, and the paper demonstrates this both by direct gradient-error measurements and by optimization results on three real-world tasks.

Load-bearing premise

The derivation assumes the black-box function is differentiable enough for the gradient theorem to apply and that paired inputs are truly close, with $\|x - x'\| \le \epsilon$; in high-dimensional problems with limited data, nearest neighbors can be far apart, and then minimizing GradPIE is not guaranteed to align the Jacobians.

Editorial extensions

If this is right

  • Training a surrogate with GradPIE instead of MAE or MSE improves gradient estimation on the same dataset, with relative error reductions over 20% and cosine-similarity gains over 8% in the oscillator experiments.
  • In active black-box optimization, the locality-aware model reaches the base model's 200-iteration performance with roughly 40-70% fewer black-box calls across the three tasks.
  • The loss works in both offline and online settings, so it can be applied when the simulator can be queried during optimization and when only a fixed dataset exists.
  • Simple MLP surrogates trained this way remain effective in a 3600-dimensional search space, indicating the method does not rely on specialized architectures.

Reading between the lines

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

  • Going beyond the paper: GradPIE could be used as an auxiliary loss or regularizer in any surrogate-training pipeline, since it only needs value differences; the paper limits its experiments to MLP surrogates and three tasks.
  • Going beyond the paper: weighting neighbor pairs by inverse distance, or annealing the neighbor radius, is a natural extension that might restore the Jacobian-alignment guarantee when data is scarce; the paper does not test this.
  • Going beyond the paper: because GradPIE ignores global offsets, it is closely related to contrastive or rank-based learning of scalar functions; that link might transfer the method to problems where only relative comparisons, not absolute values, are observed.
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 a surrogate-model training loss, GradPIE, for gradient-based black-box optimization. Starting from the gradient theorem, the authors replace the black-box function inside a path integral with the surrogate and obtain a pair-loss objective (Eq. 5), then make it scalable by using k-nearest-neighbor pairs (Eq. 10). They claim that minimizing this loss aligns the Jacobian of the surrogate with the Jacobian of the black-box function, and they evaluate the method in offline and online active black-box optimization on three tasks: coupled nonlinear oscillators (CNON), an analog OpAmp circuit, and a 3600-dimensional optical wave manipulation system (OWMS). The experiments report consistent improvements over a base surrogate trained with MAE and over Bayesian optimization and random search under limited query budgets.

Significance. If the Jacobian-alignment claim were established, GradPIE would be a simple and cheap modification to surrogate training for black-box optimization, and the paper would make a useful practical contribution. The work has real strengths: the loss is derived from an integral identity rather than from fitted constants; the CNON experiment directly compares estimated gradients with exact gradients; and the empirical evaluation spans three heterogeneous tasks, including a high-dimensional optical system. However, the theoretical relation is established only for an ideal pair loss, not for the k-NN loss that is actually optimized, and the high-dimensional experiments are conducted in a regime where the small-displacement assumption is questionable. The empirical gains are promising but, without an ablation or a formal bound, the paper's central mechanism remains conditional.

major comments (3)
  1. [Sections 3-4, Eqs. (5), (9), (10)] The Jacobian-alignment conclusion of Eq. (9) applies to the ideal pair loss in Eq. (5), but the optimized loss in Eq. (10) is a k-NN empirical average, and the paper provides no bound linking the two objectives. For a fixed x, the per-point term in Eq. (10) penalizes A = J[Fhat] - J[F] only on the K displacement vectors v_k = x'_k - x; any component of A orthogonal to span{v_1, ..., v_K} is invisible to the loss. In the OWMS task (D = 3600, N = 1000), K is necessarily much smaller than D, so the loss can be arbitrarily small while the row-norm Jacobian discrepancy is large in most directions. The claimed equivalence in the abstract and Section 3 is therefore not inherited by the scalable loss actually used in training.
  2. [Section 3, Eq. (8)] The derivation of Eq. (9) assumes paired inputs are close, with ||x - x'|| <= epsilon, and replaces the line integral of A(u) by A(x)(x' - x). The resulting O(epsilon^2) Taylor error is uncontrolled. In the OWMS experiment, 1000 training points in a 3600-dimensional space are sparse, so typical nearest-neighbor distances are not small, and the paper gives no evidence that the k-NN pairs used in Eq. (10) satisfy the epsilon condition. Without reporting pairwise distances or bounding second derivatives of F and Fhat, the theoretical justification does not apply to the high-dimensional regime on which the paper's main empirical claim rests.
  3. [Section 5, Algorithm 1, Table 1] The reported optimization gains are not uniquely attributable to GradPIE. The online procedure in Algorithm 1 simultaneously introduces local sampling (Ns), rank selection, and periodic retraining; for OpAmp and OWMS, no Ns = 0 ablation is reported, so the improvement relative to the base model could be driven by the local-sampling schedule or by the network's inductive bias rather than by Jacobian alignment. Since the theoretical mechanism is in question, an ablation that isolates the loss function under identical sampling schedules, or per-iteration gradient-alignment error on all tasks, is needed to support the paper's central claim.
minor comments (5)
  1. [Section 5, Table 3] The nearest-neighbor count K is a central hyperparameter, but the paper does not report the K values used for the OpAmp and OWMS tasks, nor the standard deviation sigma and Ns used in local sampling; please add a hyperparameter table for reproducibility.
  2. [Algorithm 1] The stopping condition 'if LGradPIE < epsilon then Break' is scale-dependent because the loss in Eq. (10) is an unnormalized sum of L1 differences; the threshold should be normalized or removed.
  3. [Figures 2f-2h] Figures 2f-2h report optimization curves without confidence intervals, and the text says the 'plots clearly show' superiority; given the stochasticity of the procedure, error bars or multiple-seed statistics are needed to support this statement.
  4. [Section 3] The statement that an MSE formulation 'would yield similar results' is unsupported; either provide an experiment or clearly label this as a conjecture.
  5. [General presentation] There are minor typographical issues, including the broken character in 'naively' and some informal usage of 'O(.)' in Table 2; these should be cleaned up.

Circularity Check

0 steps flagged · score 0.0 of 10

No circularity found: the GradPIE objective is derived from the gradient theorem, not from fitted constants; the empirical comparisons use independent benchmarks, so the central claim is not forced by construction.

full rationale

The derivation chain is self-contained. Eq. (4) defines the loss as the mismatch between the path integral of the surrogate gradient and the true function-value difference; Eq. (5) is obtained by applying the gradient theorem to the surrogate, and Eqs. (7)-(9) then show that, for close pairs, the loss is proportional to the row norm of the Jacobian difference. This is a genuine mathematical reduction rather than a definitional equivalence: the loss is defined in terms of paired outputs, and the Jacobian-alignment statement is derived from it under stated smoothness assumptions. The scalable loss in Eq. (10) is an empirical k-NN version of Eq. (5), and one may worry whether it inherits the guarantee for K << D, but that is a gap in proof or approximation quality, not a circular step. The experiments on CNON, OpAmp, and OWMS compare against exact gradients, random search, and Bayesian optimization, and the loss is not fitted to gradient labels, so the evaluation is not a fitted-input-called-prediction. The only self-citations (Momeni et al., 2023a,b) appear in contextual remarks in the impact statement and OWMS task description and do not provide any load-bearing premise for the GradPIE derivation. No self-citation chain or imported uniqueness theorem is used.

Assumptions & free parameters 4 free parameters · 3 assumptions · 0 invented entities

The central claim rests on standard vector-calculus identities plus regularity assumptions on the black-box function. The algorithm introduces tunable locality hyperparameters but no new physical or mathematical entities.

free parameters (4)
  • Number of nearest neighbors K = not reported per task
    Controls the locality of GradPIE in Eq. (10); the authors state optimal K is case-dependent in Section 5.1, and values are not given.
  • Local sampling standard deviation sigma = not reported
    Used in Algorithm 1 to generate local samples around the updated input; the value is not specified.
  • Number of local samples Ns = 0 or 1 for CNON, unspecified elsewhere
    Controls online data augmentation in Algorithm 1; Table 1 reports Ns=0 and Ns=1 for CNON, with no values or tuning procedure for the other tasks.
  • Training hyperparameters (epochs, learning rates, batch size, Nbest, epsilon) = not reported
    Algorithm 1 lists these as inputs without concrete values, affecting reproducibility rather than the conceptual claim.
assumptions (3)
  • domain assumption The black-box function F is differentiable along integration paths so the gradient theorem applies.
    Section 3 uses Eq. (3) on F even though the problem setup in Section 2 highlights non-differentiable simulators.
  • domain assumption Paired inputs are close enough for the local linearization in Eq. (8): the integral of the Jacobian difference over the path is approximated by Jdiff(x)(x-prime minus x).
    Used to transform Eq. (7) into Eq. (9); assumes the distance between x and x-prime is at most epsilon and that p(x, x-prime) is continuous.
  • ad hoc to paper K-nearest-neighbor pairs provide a representative sample of the local neighborhoods needed for Jacobian alignment.
    Eq. (10) replaces the theoretical expectation over close pairs with an average over K nearest neighbors; no justification is given that KNN distances fall within the local regime in high dimensions.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Locality-aware Surrogates for Gradient-based Black-box Optimization." pith.science (2026). https://pith.science/paper/GL36STDV

@misc{pith2026250119161,
  author       = {Pith},
  title        = {Pith review of: Locality-aware Surrogates for Gradient-based Black-box Optimization},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/GL36STDV}},
  note         = {Machine review of arXiv:2501.19161}
}
read the original abstract

In physics and engineering, many processes are modeled using non-differentiable black-box simulators, making the optimization of such functions particularly challenging. To address such cases, inspired by the Gradient Theorem, we propose locality-aware surrogate models for active model-based black-box optimization. We first establish a theoretical connection between gradient alignment and the minimization of a Gradient Path Integral Equation (GradPIE) loss, which enforces consistency of the surrogate's gradients in local regions of the design space. Leveraging this theoretical insight, we develop a scalable training algorithm that minimizes the GradPIE loss, enabling both offline and online learning while maintaining computational efficiency. We evaluate our approach on three real-world tasks - spanning automated in silico experiments such as coupled nonlinear oscillators, analog circuits, and optical systems - and demonstrate consistent improvements in optimization efficiency under limited query budgets. Our results offer dependable solutions for both offline and online optimization tasks where reliable gradient estimation is needed.

Figures

Figures reproduced from arXiv: 2501.19161 by the authors.

Figure 1
Figure 1. a Our method calculates the k-nearest neighbors for each sample in the dataset. These neighbors are used to train the locality￾aware surrogate model with the GradPIE loss function. b Active black-box optimization (ABBO) using the surrogate model trained offline. c ABBO with online training of the surrogate model. optimization problems where automated querying of the black-box function is possible, typically in autom… view at source ↗
Figure 2
Figure 2. Relative error and cosine similarity between estimated and exact gradient for input dimensions of Di = 7 (a and b) and Di = 10 (c and d ), as a function of the number of nearest neighbors for CNON. e The improvement in gradient estimation for optimal number of nearest neighbors. Performance of ABBO using the offline-trained surrogate model for the CNON task for f λ = 0.50, g λ = 0.55, and h λ = 0.70. rewritten as LG… view at source ↗
Figure 3
Figure 3. Performance of ABBO using the online￾trained surrogate model for the CNON task (λ = [−0.55, 0.125, 0.31, −0.38, 0.60]): a without local sam￾pling (Ns = 0) and b with local sampling (Ns = 1). c Schematic of the OpAmp circuit task. d Mean ± standard deviation and e median with interquartile range (IQR) of ABBO using the online-trained surrogate model for the OpAmp task. Here, the impulse response hz is expressed as hz… view at source ↗
Figures from the paper (1 more)
Figure 4
Figure 4. Figure 4: a Schematic of the OWMS. An input Gaussian beam is illuminated onto a Spatial Light Modulator (SLM) with 3600 parameters to be optimized such that the output waveform matches the target waveform. b Mean ± standard deviation and c median with interquartile range (IQR) o…

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

48 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]

    J., Lu, W., Chang, M.-F., Ielmini, D., Yang, Y., et al

    Aguirre, F., Sebastian, A., Le Gallo, M., Song, W., Wang, T., Yang, J. J., Lu, W., Chang, M.-F., Ielmini, D., Yang, Y., et al. Hardware implementation of memristor-based artificial neural networks. Nature communications, 15 0 (1): 0 1974, 2024

    work page 1974

  3. [3]

    Multi-objective optimization in material design and selection

    Ashby, M. Multi-objective optimization in material design and selection. Acta materialia, 48 0 (1): 0 359--369, 2000

    work page 2000

  4. [4]

    E., and Francone, F

    Banzhaf, W., Nordin, P., Keller, R. E., and Francone, F. D. Genetic programming: an introduction: on the automatic evolution of computer programs and its applications. Morgan Kaufmann Publishers Inc., 1998

    work page 1998

  5. [5]

    Single-shot optical neural network

    Bernstein, L., Sludds, A., Panuski, C., Trajtenberg-Mills, S., Hamerly, R., and Englund, D. Single-shot optical neural network. Science Advances, 9 0 (25): 0 eadg7904, 2023

    work page 2023

  6. [6]

    Braun, O. M. and Kivshar, Y. S. Nonlinear dynamics of the frenkel--kontorova model. Physics Reports, 306 0 (1-2): 0 1--108, 1998

    work page 1998

  7. [7]

    Conditioning by adaptive sampling for robust design

    Brookes, D., Park, H., and Listgarten, J. Conditioning by adaptive sampling for robust design. In International conference on machine learning, pp.\ 773--782. PMLR, 2019

    2019

  8. [8]

    N., and Doppa, J

    Chemingui, Y., Deshwal, A., Hoang, T. N., and Doppa, J. R. Offline model-based optimization via policy-guided gradient search. In Proceedings of the AAAI Conference on Artificial Intelligence, volume 38, pp.\ 11230--11239, 2024

    work page 2024

Show all 48 references
  1. [9]

    C., Le Nguyen, P., Truong, T

    Dao, M. C., Le Nguyen, P., Truong, T. N., and Hoang, T. N. Boosting offline optimizers with surrogate sensitivity. In Forty-first International Conference on Machine Learning, 2024

    2024

  2. [10]

    de Avila Belbute-Peres, F., Smith, K., Allen, K., Tenenbaum, J., and Kolter, J. Z. End-to-end differentiable physics for learning and control. Advances in neural information processing systems, 31, 2018

    2018

  3. [11]

    A differentiable physics engine for deep learning in robotics

    Degrave, J., Hermans, M., Dambre, J., and Wyffels, F. A differentiable physics engine for deep learning in robotics. Frontiers in neurorobotics, 13: 0 6, 2019

    2019

  4. [12]

    Cktgnn: Circuit graph neural network for electronic design automation

    Dong, Z., Cao, W., Zhang, M., Tao, D., Chen, Y., and Zhang, X. Cktgnn: Circuit graph neural network for electronic design automation. arXiv preprint arXiv:2308.16406, 2023

    2023 arXiv

  5. [13]

    D., and Poloczek, M

    Eriksson, D., Pearce, M., Gardner, J., Turner, R. D., and Poloczek, M. Scalable global optimization via local bayesian optimization. Advances in neural information processing systems, 32, 2019

    2019

  6. [14]

    and Listgarten, J

    Fannjiang, C. and Listgarten, J. Autofocused oracles for model-based design. Advances in Neural Information Processing Systems, 33: 0 12945--12956, 2020

    2020

  7. [15]

    Filipovich, M. J. and Lvovsky, A. Torchoptics: An open-source python library for differentiable fourier optics simulations. arXiv preprint arXiv:2411.18591, 2024

    2024 arXiv

  8. [16]

    and Levine, S

    Fu, J. and Levine, S. Offline model-based optimization via normalized maximum likelihood estimation. arXiv preprint arXiv:2102.07970, 2021

    2021 arXiv

  9. [17]

    Backpropagation through the void: Optimizing control variates for black-box gradient estimation

    Grathwohl, W., Choi, D., Wu, Y., Roeder, G., and Duvenaud, D. Backpropagation through the void: Optimizing control variates for black-box gradient estimation. arXiv preprint arXiv:1711.00123, 2017

    2017 arXiv

  10. [18]

    H., and Leyton-Brown, K

    Hutter, F., Hoos, H. H., and Leyton-Brown, K. Sequential model-based optimization for general algorithm configuration. In Learning and Intelligent Optimization: 5th International Conference, LION 5, Rome, Italy, January 17-21, 2011. Selected Papers 5, pp.\ 507--523. Springer, 2011

    2011

  11. [19]

    L., Davenport, J

    Iman, R. L., Davenport, J. M., and Zeigler, D. K. Latin hypercube sampling (program user's guide).[lhc, in fortran]. Technical report, Sandia Labs., Albuquerque, NM (USA), 1980

    1980

  12. [20]

    M., and Grover, A

    Krishnamoorthy, S., Mashkaria, S. M., and Grover, A. Generative pretraining for black-box optimization. arXiv preprint arXiv:2206.10786, 2022

    2022 arXiv

  13. [21]

    and Levine, S

    Kumar, A. and Levine, S. Model inversion networks for model-based optimization. Advances in neural information processing systems, 33: 0 5126--5137, 2020

    2020

  14. [22]

    K., and Mishra, S

    Lanthaler, S., Rusch, T. K., and Mishra, S. Neural oscillators are universal. Advances in Neural Information Processing Systems, 36, 2024

    2024

  15. [23]

    Adversarial variational optimization of non-differentiable simulators

    Louppe, G., Hermans, J., and Cranmer, K. Adversarial variational optimization of non-differentiable simulators. In The 22nd International Conference on Artificial Intelligence and Statistics, pp.\ 1438--1447. PMLR, 2019

    2019

  16. [24]

    G., and McMahon, P

    Ma, S.-Y., Wang, T., Laydevant, J., Wright, L. G., and McMahon, P. L. Quantum-limited stochastic optical neural networks operating at a few quanta per activation. Nature Communications, 16 0 (1): 0 359, 2025

    2025

  17. [25]

    McMahon, P. L. The physics of optical computing. Nature Reviews Physics, 5 0 (12): 0 717--734, 2023

    2023

  18. [26]

    K., Molina-Mendoza, A

    Mennel, L., Symonowicz, J., Wachter, S., Polyushkin, D. K., Molina-Mendoza, A. J., and Mueller, T. Ultrafast machine vision with 2d material neural network image sensors. Nature, 579 0 (7797): 0 62--66, 2020

    2020

  19. [27]

    Monte carlo gradient estimation in machine learning

    Mohamed, S., Rosca, M., Figurnov, M., and Mnih, A. Monte carlo gradient estimation in machine learning. Journal of Machine Learning Research, 21 0 (132): 0 1--62, 2020

    2020

  20. [28]

    Backpropagation-free training of deep physical neural networks

    Momeni, A., Rahmani, B., Mall \'e jac, M., Del Hougne, P., and Fleury, R. Backpropagation-free training of deep physical neural networks. Science, 382 0 (6676): 0 1297--1303, 2023 a

    2023

  21. [29]

    Phyff: Physical forward forward algorithm for in-hardware training and inference

    Momeni, A., Rahmani, B., Mall \'e jac, M., del Hougne, P., and Fleury, R. Phyff: Physical forward forward algorithm for in-hardware training and inference. In Machine Learning with New Compute Paradigms, 2023 b

    2023

  22. [30]

    Nguyen, A. W. and Daugherty, P. S. Evolutionary optimization of fluorescent proteins for intracellular fret. Nature biotechnology, 23 0 (3): 0 355--360, 2005

    2005

  23. [31]

    U., Yildirim, M., Ke, J., Yoo, I., Wang, Q., Yang, F., Moser, C., and Psaltis, D

    Oguz, I., Dinc, N. U., Yildirim, M., Ke, J., Yoo, I., Wang, Q., Yang, F., Moser, C., and Psaltis, D. Optical diffusion models for image generation. arXiv preprint arXiv:2407.10897, 2024

    2024 arXiv

  24. [32]

    Learning to simulate

    Ruiz, N., Schulter, S., and Chandraker, M. Learning to simulate. arXiv preprint arXiv:1810.02513, 2018

    2018 arXiv

  25. [33]

    S., Bolotin, D

    Sarkisyan, K. S., Bolotin, D. A., Meer, M. V., Usmanova, D. R., Mishin, A. S., Sharonov, G. V., Ivankov, D. N., Bozhanova, N. G., Baranov, M. S., Soylemez, O., et al. Local fitness landscape of the green fluorescent protein. Nature, 533 0 (7603): 0 397--401, 2016

    2016

  26. [34]

    P., and De Freitas, N

    Shahriari, B., Swersky, K., Wang, Z., Adams, R. P., and De Freitas, N. Taking the human out of the loop: A review of bayesian optimization. Proceedings of the IEEE, 104 0 (1): 0 148--175, 2015

    2015

  27. [35]

    Shirobokov, S., Belavin, V., Kagan, M., Ustyuzhanin, A., and Baydin, A. G. Black-box optimization with local generative surrogates. Advances in neural information processing systems, 33: 0 14650--14662, 2020

    2020

  28. [36]

    High-temperature superconductivity in iron pnictides and chalcogenides

    Si, Q., Yu, R., and Abrahams, E. High-temperature superconductivity in iron pnictides and chalcogenides. Nature Reviews Materials, 1 0 (4): 0 1--15, 2016

    2016

  29. [37]

    Snoek, J., Larochelle, H., and Adams, R. P. Practical bayesian optimization of machine learning algorithms. Advances in neural information processing systems, 25, 2012

    2012

  30. [38]

    Spall, J., Guo, X., and Lvovsky, A. I. Hybrid training of optical neural networks. Optica, 9 0 (7): 0 803--811, 2022

    2022

  31. [39]

    The method of moving asymptotes—a new method for structural optimization

    Svanberg, K. The method of moving asymptotes—a new method for structural optimization. International journal for numerical methods in engineering, 24 0 (2): 0 359--373, 1987

    1987

  32. [40]

    Conservative objective models for effective offline model-based optimization

    Trabucco, B., Kumar, A., Geng, X., and Levine, S. Conservative objective models for effective offline model-based optimization. In International Conference on Machine Learning, pp.\ 10358--10368. PMLR, 2021

    2021

  33. [41]

    Stochastic zeroth-order optimization in high dimensions

    Wang, Y., Du, S., Balakrishnan, S., and Singh, A. Stochastic zeroth-order optimization in high dimensions. In International conference on artificial intelligence and statistics, pp.\ 1356--1365. PMLR, 2018

    2018

  34. [42]

    Williams, R. J. Simple statistical gradient-following algorithms for connectionist reinforcement learning. Machine learning, 8: 0 229--256, 1992

    1992

  35. [43]

    Williamson, R. E. and Trotter, H. F. Multivariable Mathematics. Pearson Prentice Hall, 4th edition, 2004

    2004

  36. [44]

    G., Onodera, T., Stein, M

    Wright, L. G., Onodera, T., Stein, M. M., Wang, T., Schachter, D. T., Hu, Z., and McMahon, P. L. Deep physical neural networks trained with backpropagation. Nature, 601 0 (7894): 0 549--555, 2022

    2022

  37. [45]

    J., and Qian, H

    Yao, P., Wu, H., Gao, B., Tang, J., Zhang, Q., Zhang, W., Yang, J. J., and Qian, H. Fully hardware-implemented memristor convolutional neural network. Nature, 577 0 (7792): 0 641--646, 2020

    2020

  38. [46]

    Roma: Robust model adaptation for offline model-based optimization

    Yu, S., Ahn, S., Song, L., and Shin, J. Roma: Robust model adaptation for offline model-based optimization. Advances in Neural Information Processing Systems, 34: 0 4619--4631, 2021

    2021

  39. [47]

    Dual adaptive training of photonic neural networks

    Zheng, Z., Duan, Z., Chen, H., Yang, R., Gao, S., Zhang, H., Xiong, H., and Lin, X. Dual adaptive training of photonic neural networks. Nature Machine Intelligence, 5 0 (10): 0 1119--1129, 2023

    2023

  40. [48]

    and Chai, Y

    Zhou, F. and Chai, Y. Near-sensor and in-sensor computing. Nature Electronics, 3 0 (11): 0 664--671, 2020

    2020

Pith tools

Reviewed August 9, 2026 · model on record in the stance chip above.