Pith. sign in

REVIEW 4 major objections 4 minor 53 references

Physics-Guided Interpretable Machine Learning Framework for Anomalous Transport in Crowded Media with Tunable Flexibility

T0 review · 4 major / 4 minor · reviewed 2026-08-01 · deepseek-v4-flash

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

Pith's one-line read Bond flexibility independently regulates post-network relaxation in crowded colloidal networks without altering network topology.

desk verdict Solid simulation/surrogate framework with a plausible stage-dependent flexibility effect, but the central dynamic mechanism is confounded by the static structural changes that p_flex itself induces. read the letter →

arxiv 2607.25827 v1 pith:TXV5WYDO submitted 2026-07-28 cond-mat.soft cond-mat.mtrl-sci

Zakiya Shireen , Sujin B. Babu This is my paper · ORCID
classification cond-mat.softcond-mat.mtrl-sci PACS 82.70.Dd
keywords anomaloustransportcolloidalnetworksbondflexibilitytransientcagingBrowniansimulationinterpretablemachinelearningrelaxationtimescrowdedmedia
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 argues that bond flexibility is a distinct, stage-dependent control parameter for anomalous transport in crowded colloidal networks, on par with crowding and composition but acting through a different mechanism. Before percolated network formation, crowding dominates relaxation; after network formation, flexibility enables local bond rearrangements that release explorer particles from transient cages, shortening relaxation times. The authors support this by combining particle-resolved Brownian simulations with surrogate machine-learning models and game-theoretic attribution to separate the contributions of total volume fraction, explorer fraction, and bond flexibility across a full parameter space. They also show that the influence of flexibility grows with explorer fraction, suggesting that network mechanics is a tunable design variable for transport in soft materials.

What carries the argument

The central object is the bond-flexibility parameter p_flex, which sets the probability that a bonded template particle undergoes Brownian displacement at each movement step while maintaining its irreversible bond. This parameter controls local network rearrangements without altering connectivity. The analysis relies on surrogate machine-learning models trained on structural and dynamical descriptors, combined with game-theoretic attribution to decompose the predicted descriptors into contributions from total volume fraction, explorer fraction, and p_flex. This setup lets the authors rank parameter importance globally and locally, and to expose conditional interactions between parameters.

What would settle it

Compute the cluster-size distribution, mean coordination number, and pore-size distribution of the template network for two states matched in total volume fraction and explorer fraction but differing in p_flex, at the same stage; if these static quantities change appreciably with p_flex, the claim that flexibility acts without altering network topology would be falsified.

Watch

Extended reading notes

Core claim

The central claim is that bond flexibility, governed by a parameter p_flex controlling the mobility of bonded template particles without breaking irreversible bonds, acts as an independent regulator of post-network relaxation. In the fully formed template network, increasing p_flex promotes local bond rearrangements that help explorer particles escape transient cages, thereby speeding up the long-time relaxation and narrowing the displacement distribution, while leaving the irreversible network topology unchanged. The authors find that the importance of flexibility is stage-dependent: negligible before network formation, substantial afterward, and most pronounced in explorer-rich systems. Th

Load-bearing premise

The p_flex protocol changes only the mobility of bonded template particles while preserving the irreversible network topology; if varying p_flex also changes connectivity, cluster-size distribution, or effective cage structure, then the relaxation changes cannot be attributed solely to bond flexibility.

Editorial extensions

If this is right

  • Flexible template networks can speed up explorer relaxation without breaking the network, offering a structural-preserving knob to tune transport in gels and porous media.
  • The effect of flexibility is conditional on explorer fraction, so designing a network with a given transport response requires knowing the particle composition as well as the mesh structure.
  • Before network formation, flexibility barely matters; after it, flexibility becomes a major relaxation lever, so tuning should be applied at the post-network stage.
  • The descriptors used (mean-squared displacement, intermediate scattering function, displacement distribution) are measurable in experiments, so the predicted importance rankings could be tested with tracer diffusion in colloidal gels.
  • The analytical workflow can be transferred to other particle-resolved simulations or experimental datasets where multiple physical variables jointly control transport.

Reading between the lines

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

  • If flexibility works through a cage-release mechanism without topology change, then simulations that freeze connectivity but allow bond rotations or vibrations should reproduce the same relaxation speed-up; this would directly test the proposed mechanism.
  • In biological contexts such as cytoplasm or polymer gels, local network fluctuations may matter as much as static mesh size for tracer diffusion; experiments with cross-linkers of different lability could validate this.
  • Because flexibility accelerates relaxation kinetically rather than thermodynamically, it may allow decoupling of transport enhancement from mechanical rigidity, which could be useful in designing self-healing or adaptive materials.
  • The stage-dependent importance of flexibility suggests that interpretation of transport data in evolving networks must account for network maturity, not just average structural descriptors.
Share X Bluesky LinkedIn Reddit HN

Signed reviews

No signed human review yet.

Editorial analysis

A structured set of objections, weighed in public.

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

Referee Report

4 major / 4 minor

Summary. The paper presents a Brownian Cluster Dynamics (BCD) simulation study of binary colloidal mixtures in which explorer particles diffuse through a template network of irreversibly bonded particles whose bond mobility is controlled by a flexibility parameter p_flex. The authors combine structural descriptors (mean minimum explorer–template separation, largest-cluster fraction), dynamical observables (MSD, intermediate scattering function, displacement distributions) with Random Forest surrogate models and SHAP attribution to separate the influence of total volume fraction Φ_tot, explorer fraction c_E, and bond flexibility p_flex, both before and after template-network formation. The central claim is that, after network formation, bond flexibility acts as an independent regulator of explorer relaxation: local bond rearrangements facilitate escape from transient cages without altering the irreversible network topology, with the effect growing with explorer fraction. The paper also proposes the workflow as a general physics-guided interpretable ML framework for disentangling coupled transport mechanisms.

Significance. If the central claim is correct, the paper identifies bond flexibility as a distinct control parameter for post-network relaxation in heterogeneous colloidal networks, going beyond crowding and composition. The study covers a broad parameter space and makes a serious effort to connect machine-learned attributions to direct physics-based comparisons; the controlled F(q,t) comparisons in Fig. 6, which hold φ_E or φ_T fixed while varying p_flex, are a particularly strong element. The machine-learning framework itself is sensible and could be valuable for other soft-matter problems. However, the central mechanistic attribution—that p_flex acts dynamically without changing the static confinement landscape—is not yet established. The paper's own structural analysis shows that p_flex changes the static pore structure, and the SHAP-based quantitative disentangling rests on a surrogate whose validation is only asserted. These gaps are load-bearing for the main conclusions and need to be addressed before the results can be considered fully supported.

major comments (4)
  1. [Section III.A, Fig. 2b; Section III.C.2, Fig. 6] The central claim that bond flexibility regulates relaxation by promoting local bond rearrangements without altering the irreversible network topology is confounded unless p_flex is shown to preserve the static confinement landscape. The text in Section III.A explicitly states that increasing p_flex 'permits local bond rearrangements that produce more open configurations and larger accessible voids,' and Fig. 2b shows that both ⟨r_min⟩ and f_max vary with p_flex at fixed Φ_tot. The controlled comparisons in Fig. 6 hold φ_E or φ_T fixed, but not the pore/cage structure, connectivity, or cluster-size distribution. Therefore, the faster relaxation at higher p_flex could be due to a less confining static network rather than dynamic rearrangement-assisted cage release. Please quantify the invariance of the static structure under p_flex at matched Φ_tot and c_E: e.g., bond graph, largest-clust
  2. [Section II.B, Section III.A.1] The p_flex protocol may also alter the final bond topology through pre-network aggregation dynamics, since bonded template particles remain mobile during cluster formation and cluster diffusion can change the aggregation pathway. Section II.B states that the protocol preserves network connectivity, but connectivity is not the same as the full static cage/pore structure. The structural descriptor maps in Fig. 2b indicate p_flex-dependent changes in r_min and f_max in the pre-network stage, which can propagate into the post-network structure. A direct test of bond-topology invariance (e.g., bond lifetime, coordination number, cluster-size distribution, or pore-size distribution) across p_flex at fixed Φ_tot and c_E is needed to rule out a structural explanation for the observed relaxation differences.
  3. [Section II.A; Figs. 3–7] The quantitative SHAP attribution rests on a Random Forest surrogate whose validation is only asserted ('The surrogate models are validated prior to the SHAP analysis'), with no reported metrics such as R², mean absolute error, or holdout performance. Since the SHAP analysis interprets the same descriptors on which the surrogate was trained, the global and local attributions are partly self-referential. Please provide quantitative validation of the surrogates for each target descriptor (α, τ_α, w_v, etc.), including train/test splits or cross-validation, and report uncertainties in the SHAP values (e.g., bootstrap or variance across trees). Without this, the quantitative disentangling claims in Figs. 5(d) and 7(c) are not fully supported.
  4. [Sections III.B–III.D] The values of α, τ_α, and w_v are presented without error bars or uncertainties, even though they are used to infer quantitative trends (e.g., α≈0.74 vs. 0.36 in Fig. 4a, or the p_flex-dependent gradients in Fig. 5b). Given the modest number of representative states and the noise inherent in finite-size simulations, error estimates (e.g., over independent realizations or block averages) are needed to establish that the reported differences, especially the subtle p_flex effects at low explorer fraction, are statistically meaningful. This is particularly important because the local SHAP and dependence analyses rely on differences that appear small compared with the dominant Φ_tot/c_E effects.
minor comments (4)
  1. [General notation] The text switches between Φ_tot, φ_E, and φ_T without a consistent definition; Fig. 6 introduces φ_E and φ_T but the relationship to c_E and Φ_tot is not explicitly stated. Define all symbols at first use and use consistently throughout.
  2. [Section II.B] The criterion for 'network growth reaches a steady state' is not quantitatively defined. Please specify the measured quantity (e.g., f_max or cluster-size distribution) and the convergence tolerance used to define the post-network stage.
  3. [Data availability] The Data Availability statement says data and scripts are available 'upon reasonable request.' Given the emphasis on a reproducible framework, the authors should deposit the simulation data, analysis scripts, and trained surrogate models in a public repository.
  4. [Title and text] Minor typographical and formatting issues include the title 'T ransport' with a stray space, 'a increasingly stronger dependence' in Section III.C.2, and several reference entries with corrupted author names (e.g., 'V olpe' in ref. 11). Please proofread.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: the relaxation finding is supported by direct simulations, and the SHAP/surrogate analysis is an interpretive tool rather than a fitted prediction.

full rationale

The paper's derivation chain is self-contained: BCD simulations generate trajectories for systematically varied (Phi_tot, c_E, p_flex); structural and dynamical descriptors are extracted; a Random Forest surrogate is trained and validated; SHAP attributes the surrogate's predictions to inputs; and the central relaxation claim is tested by direct controlled F(q,t) comparisons in Fig. 6 while holding phi_E or phi_T fixed. No equation or result is defined in terms of the quantity it is said to predict. The SHAP analysis is an interpretation of a fitted surrogate, not an out-of-sample prediction, and the paper does not treat SHAP importances as a substitute for the direct simulations; it uses SHAP to select representative states and to summarize learned trends. The p_flex protocol is cited to the authors' prior work (ref. 43), but the protocol's connectivity-preserving property is part of the simulation method, and the relaxation finding rests on the present simulations rather than on the citation. The potential confound that p_flex also changes static void/cage structure (Fig. 2b) is a threat to causal attribution but is a correctness/experimental-design issue, not circularity: the observed relaxation differences are not equivalent to the protocol definition by construction. Therefore no circular step is identifiable under the required standard.

Assumptions & free parameters 1 free parameters · 4 assumptions · 0 invented entities

No new physical entities are introduced. The central claim rests on the BCD model, the p_flex protocol, and the RF+SHAP attribution assumptions; the latter are the least externally validated.

free parameters (1)
  • Random Forest hyperparameters (number of trees, depth, split criteria, etc.) = not reported
    The surrogate and all SHAP rankings depend on these chosen values; without them, the quantitative attribution cannot be independently reproduced or assessed.
assumptions (4)
  • domain assumption BCD algorithm with irreversible bond formation faithfully simulates binary colloidal gelation
    Invoked in Sec. II B, based on refs 19,39,41,42; if the model misses relevant physics (e.g., hydrodynamic interactions), the quantitative attributions could shift.
  • domain assumption p_flex changes only bonded-template mobility and preserves irreversible network connectivity/topology
    Introduced in Sec. II B and attributed to ref 43; central to interpreting flexibility as an independent regulator. No topology invariance check is shown.
  • ad hoc to paper Random Forest surrogate reliably approximates descriptor dependence and SHAP values quantify physical contributions
    Assumed in Sec. II A; validation is asserted but metrics/hyperparameters/train-test details are not reported.
  • domain assumption Descriptors from 200 realizations and selected q/time points are converged
    Used in Sec. II A for MSD, F(q,t), and w_v; no convergence analysis is shown.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Physics-Guided Interpretable Machine Learning Framework for Anomalous Transport in Crowded Media with Tunable Flexibility." pith.science (2026). https://pith.science/paper/TXV5WYDO

@misc{pith2026260725827,
  author       = {Pith},
  title        = {Pith review of: Physics-Guided Interpretable Machine Learning Framework for Anomalous Transport in Crowded Media with Tunable Flexibility},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/TXV5WYDO}},
  note         = {Machine review of arXiv:2607.25827}
}
read the original abstract

Transport in crowded media is governed by the interplay of multiple physical mechanisms. Quantitatively disentangling their individual and coupled contributions remains a longstanding challenge because the evolving microstructure continuously modifies their relative influence. Here, we develop a physics-guided interpretable machine-learning framework that couples Brownian Cluster Dynamics simulations with surrogate machine-learning models and SHAP-based interpretation to quantitatively disentangle the individual and coupled effects of total volume fraction, explorer fraction, and template bond flexibility on explorer-particle transport. We demonstrate the framework using binary colloidal systems in which explorer particles diffuse through a template network formed by irreversible bonds with tunable flexibility. Structural descriptors, mean-squared displacement, intermediate scattering functions, and displacement distributions reveal that network formation localizes explorer particles through enhanced transient caging. Bond flexibility emerges as an independent regulator of post-network relaxation by promoting local bond rearrangements that facilitate the release of transiently caged particles without altering the irreversible network topology. Although crowding and composition dominate the overall transport response, quantitative attribution reveals how the relative contributions of crowding, composition, and template bond flexibility evolve as the confining environment develops. Beyond establishing bond flexibility as a distinct control parameter for relaxation in heterogeneous colloidal networks, this work provides a general physics-guided interpretable machine-learning framework for quantitatively disentangling coupled physical mechanisms in complex transport phenomena.

Figures

Figures reproduced from arXiv: 2607.25827 by the authors.

Figure 1
Figure 1. FIG. 1 [PITH_FULL_IMAGE:figures/full_fig_p003_1.png] view at source ↗
Figure 2
Figure 2. FIG. 2 [PITH_FULL_IMAGE:figures/full_fig_p005_2.png] view at source ↗
Figure 3
Figure 3. b presents the local SHAP analysis for three rep￾resentative confinement environments. These correspond to systems characterized by high bond flexibility and low total volume fraction (high/low), intermediate values of both pa￾rameters (mid/mid), and low bond flexibility with high total volume fraction (low/high), spanning the low-, intermediate- , and high-confinement regimes, respectively. The represen￾tative syst… view at source ↗
Figures from the paper (4 more)
Figure 4
Figure 4. Figure 4: FIG. 4 [PITH_FULL_IMAGE:figures/full_fig_p008_4.png]
Figure 5
Figure 5. Figure 5: b extends the descriptor-level analysis by recon￾structing the relaxation landscape learned from the full simu￾lation datasets. The relaxation time, τα, is mapped over the full parameter space pre- and post- template network forma￾tion. Each column corresponds to a fix…
Figure 6
Figure 6. Figure 6: FIG. 6 [PITH_FULL_IMAGE:figures/full_fig_p011_6.png]
Figure 7
Figure 7. Figure 7: FIG. 7 [PITH_FULL_IMAGE:figures/full_fig_p012_7.png]

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

53 extracted references · 1 linked inside Pith

  1. [1]

    Physics reports , volume=

    The random walk's guide to anomalous diffusion: a fractional dynamics approach , author=. Physics reports , volume=. 2000 , publisher=

    2000

  2. [2]

    Reports on Progress in Physics , volume=

    Anomalous transport in the crowded world of biological cells , author=. Reports on Progress in Physics , volume=. 2013 , publisher=

    2013

  3. [3]

    Biophysical journal , volume=

    Anomalous diffusion of single particles in cytoplasm , author=. Biophysical journal , volume=. 2013 , publisher=

    2013

  4. [4]

    Biophysical journal , volume=

    Anomalous diffusion of proteins due to molecular crowding , author=. Biophysical journal , volume=. 2005 , publisher=

    2005

  5. [5]

    Biophysical journal , volume=

    Diffusion of particles in the extracellular matrix: the effect of repulsive electrostatic interactions , author=. Biophysical journal , volume=. 2010 , publisher=

    2010

  6. [6]

    ACS biomaterials science & engineering , volume=

    Untangling mucosal drug delivery: engineering, designing, and testing nanoparticles to overcome the mucus barrier , author=. ACS biomaterials science & engineering , volume=. 2022 , publisher=

    2022

  7. [7]

    Proceedings of the National Academy of Sciences , volume=

    Single-particle diffusional fingerprinting: A machine-learning framework for quantitative analysis of heterogeneous diffusion , author=. Proceedings of the National Academy of Sciences , volume=. 2021 , publisher=

    2021

  8. [8]

    The journal of physical chemistry letters , volume=

    Machine-learning solutions for the analysis of single-particle diffusion trajectories , author=. The journal of physical chemistry letters , volume=. 2023 , publisher=

    2023

Show all 53 references
  1. [9]

    New Journal of Physics , volume=

    Collective dynamics effect transient subdiffusion of inert tracers in flexible gel networks , author=. New Journal of Physics , volume=. 2014 , publisher=

    2014

  2. [10]

    Journal of Physics: Condensed Matter , volume=

    Negative differential mobility and trapping in active matter systems , author=. Journal of Physics: Condensed Matter , volume=. 2017 , publisher=

    2017

  3. [11]

    Nature , volume=

    First-passage times in complex scale-invariant media , author=. Nature , volume=. 2007 , publisher=

    2007

  4. [12]

    Nature communications , volume=

    Objective comparison of methods to decode anomalous diffusion , author=. Nature communications , volume=. 2021 , publisher=

    2021

  5. [13]

    The European Physical Journal Plus , volume=

    Machine learning analysis of anomalous diffusion , author=. The European Physical Journal Plus , volume=. 2025 , publisher=

    2025

  6. [14]

    Biophysical journal , volume=

    Single-particle diffusion characterization by deep learning , author=. Biophysical journal , volume=. 2019 , publisher=

    2019

  7. [15]

    Nature computational science , volume=

    Reliable deep learning in anomalous diffusion against out-of-distribution dynamics , author=. Nature computational science , volume=. 2024 , publisher=

    2024

  8. [16]

    The Journal of Chemical Physics , volume=

    Dynamics of equilibrium-linked colloidal networks , author=. The Journal of Chemical Physics , volume=. 2022 , publisher=

    2022

  9. [17]

    Physical Review E , volume=

    Anomalous diffusion of active Brownian particles in responsive elastic gels: Nonergodicity, non-Gaussianity, and distributions of trapping times , author=. Physical Review E , volume=. 2024 , publisher=

    2024

  10. [18]

    Reviews of modern physics , volume=

    Active particles in complex and crowded environments , author=. Reviews of modern physics , volume=. 2016 , publisher=

    2016

  11. [19]

    Physical review letters , volume=

    Particle tracking microrheology of complex fluids , author=. Physical review letters , volume=. 1997 , publisher=

    1997

  12. [20]

    ACS nano , volume=

    Enhanced heterogeneous diffusion of nanoparticles in semiflexible networks , author=. ACS nano , volume=. 2021 , publisher=

    2021

  13. [21]

    The Journal of Chemical Physics , volume=

    Network confinement and heterogeneity slows nanoparticle diffusion in polymer gels , author=. The Journal of Chemical Physics , volume=. 2017 , publisher=

    2017

  14. [22]

    National Science Review , volume=

    Reconfiguring confined magnetic colloids with tunable fluid transport behavior , author=. National Science Review , volume=. 2021 , publisher=

    2021

  15. [23]

    The Journal of Physical Chemistry B , volume=

    Macromolecular crowding: how shape and interactions affect diffusion , author=. The Journal of Physical Chemistry B , volume=. 2020 , publisher=

    2020

  16. [24]

    Europhysics Letters , volume=

    Diffusion in crowded colloids of particles cyclically changing their shapes , author=. Europhysics Letters , volume=. 2020 , publisher=

    2020

  17. [25]

    The Journal of Physical Chemistry B , volume=

    Tracer diffusion in colloidal gels , author=. The Journal of Physical Chemistry B , volume=. 2008 , publisher=

    2008

  18. [26]

    Physical Review E , volume=

    Enhancement in the diffusivity of Brownian spheroids in the presence of spheres , author=. Physical Review E , volume=. 2022 , publisher=

    2022

  19. [27]

    The Journal of Chemical Physics , volume=

    Dimensional confinement and superdiffusive rotational motion of uniaxial colloids in the presence of cylindrical obstacles , author=. The Journal of Chemical Physics , volume=. 2025 , publisher=

    2025

  20. [28]

    Science advances , volume=

    First-passage statistics of colloids on fractals: Theory and experimental realization , author=. Science advances , volume=. 2022 , publisher=

    2022

  21. [29]

    The Journal of Chemical Physics , volume=

    Tracer dynamics in polymer networks: Generalized Langevin description , author=. The Journal of Chemical Physics , volume=. 2024 , publisher=

    2024

  22. [30]

    Advances in neural information processing systems , volume=

    A unified approach to interpreting model predictions , author=. Advances in neural information processing systems , volume=

  23. [31]

    arXiv preprint arXiv:1802.03888 , year=

    Consistent individualized feature attribution for tree ensembles , author=. arXiv preprint arXiv:1802.03888 , year=

    arXiv

  24. [32]

    Chemical Science , volume=

    Polymer design via SHAP and Bayesian machine learning optimizes pDNA and CRISPR ribonucleoprotein delivery , author=. Chemical Science , volume=. 2024 , publisher=

    2024

  25. [33]

    Soft Matter , volume=

    Morphology prediction for polymer blend thin films using machine learning , author=. Soft Matter , volume=. 2025 , publisher=

    2025

  26. [34]

    Journal of Fluid Mechanics , volume=

    Interpreted machine learning in fluid dynamics: explaining relaminarisation events in wall-bounded shear flows , author=. Journal of Fluid Mechanics , volume=. 2022 , publisher=

    2022

  27. [35]

    Machine learning , volume=

    Random forests , author=. Machine learning , volume=. 2001 , publisher=

    2001

  28. [36]

    Testing mode-coupling theory for a supercooled binary Lennard-Jones mixture. II. Intermediate scattering function and dynamic susceptibility , author=. Physical Review E , volume=. 1995 , publisher=

    1995

  29. [37]

    Physical Review E , volume=

    Relaxation dynamics of a viscous silica melt: The intermediate scattering functions , author=. Physical Review E , volume=. 2001 , publisher=

    2001

  30. [38]

    Physical Review Letters , volume=

    Formation of fractal clusters and networks by irreversible diffusion-limited aggregation , author=. Physical Review Letters , volume=. 1983 , publisher=

    1983

  31. [39]

    Physical Review Letters , volume=

    Scaling of kinetically growing clusters , author=. Physical Review Letters , volume=. 1983 , publisher=

    1983

  32. [40]

    The Journal of chemical physics , volume=

    The influence of bond rigidity and cluster diffusion on the self-diffusion of hard spheres with square well interaction , author=. The Journal of chemical physics , volume=. 2008 , publisher=

    2008

  33. [41]

    Proceedings of the National Academy of Sciences , volume=

    Network physics of attractive colloidal gels: Resilience, rigidity, and phase diagram , author=. Proceedings of the National Academy of Sciences , volume=. 2024 , publisher=

    2024

  34. [42]

    Proceedings of the National Academy of Sciences , volume=

    Structural origin of relaxation in dense colloidal suspensions , author=. Proceedings of the National Academy of Sciences , volume=. 2024 , publisher=

    2024

  35. [43]

    Reports on progress in Physics , volume=

    Relaxation processes in supercooled liquids , author=. Reports on progress in Physics , volume=

  36. [44]

    Soft Matter , volume=

    Small particle dynamics in glassy polymers: diffusion, relaxation, and machine-learned softness , author=. Soft Matter , volume=. 2026 , publisher=

    2026

  37. [45]

    Soft Matter , volume=

    Characterizing dynamic heterogeneities during nanogel degradation , author=. Soft Matter , volume=. 2025 , publisher=

    2025

  38. [46]

    The Journal of Chemical Physics , volume=

    Uncertainty-aware Bayesian inference of glass transition temperatures from molecular simulations , author=. The Journal of Chemical Physics , volume=. 2026 , publisher=

    2026

  39. [47]

    Macromolecular Theory and Simulations , volume=

    Bayesian Optimization in Polymer Modeling: From Coarse-Graining Foundations to Autonomous Inverse Design , author=. Macromolecular Theory and Simulations , volume=. 2026 , publisher=

    2026

  40. [48]

    Computational Materials Science , volume=

    Interpretable surrogate modeling for polymers: A SHAP-guided framework for descriptor-level analysis , author=. Computational Materials Science , volume=. 2026 , publisher=

    2026

  41. [49]

    The Journal of Chemical Physics , volume=

    Lattice animals in diffusion limited binary colloidal system , author=. The Journal of Chemical Physics , volume=. 2017 , publisher=

    2017

  42. [50]

    Soft matter , volume=

    Cage dynamics leads to double relaxation of the intermediate scattering function in a binary colloidal system , author=. Soft matter , volume=. 2018 , publisher=

    2018

  43. [51]

    ACS omega , volume=

    Rigidity-Induced Controlled Aggregation of Binary Colloids , author=. ACS omega , volume=. 2023 , publisher=

    2023

  44. [52]

    The Journal of Chemical Physics , volume=

    Molecular transport in liquids and glasses , author=. The Journal of Chemical Physics , volume=. 1959 , publisher=

    1959

  45. [53]

    Science , volume=

    Three-dimensional direct imaging of structural relaxation near the colloidal glass transition , author=. Science , volume=. 2000 , publisher=

    2000

Pith tools

Reviewed August 1, 2026 · model on record in the stance chip above.