{"id":"581fbeaa-9b89-4f5c-acc7-feb9ffd212c5","arxiv_id":"2510.13199","paper_version":3,"verdict":"UNVERDICTED","confidence":"LOW","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":2,"one_line_summary":"NSIPF represents density via empirical particle measures and the field via a CNN trained on synthetic data, preserving mass and nonnegativity while simulating 3D multi-bump chemotaxis dynamics faster than finite difference or standard SIPF methods.","lead":"The paper presents a neural stochastic interacting particle-field (NSIPF) algorithm for solving a parabolic-hyperbolic Keller-Segel chemotaxis system in 3D, using particles to represent cell density and a convolutional neural network for the chemoattractant field. This hybrid mesh-free method aims to handle sharp gradients more efficiently than grid-based solvers for applications like tumor angiogenesis modeling.","discovery_kind":"new_method","skeptic_critique":{"model":"grok-4.3","headline":"CNN trained on synthetic data may not generalize accurately enough to drive correct 3D multi-bump particle dynamics","rationale":"The reader's weakest assumption correctly isolates the single approximation step whose failure would invalidate the claim of correctly capturing the dynamics at higher speed. All other properties (mass/nonnegativity preservation) follow from the particle discretization independently of the field solver, so the load-bearing risk is localized to CNN generalization.","tokens_in":1754,"tokens_out":329,"duration_ms":17241,"concrete_test":"Generate a reference 3D multi-bump trajectory using a high-resolution SIPF solver on the regularized system; feed the same particle positions into the trained CNN at multiple time steps and compute the L2 and max-norm error of the predicted field against the reference field; if either error exceeds 5-10% in regions of high gradient, re-run the NSIPF particles with the reference field and check whether the final bump locations or aggregation times differ by more than 10%.","verdict_should_be":"CONDITIONAL","load_bearing_attack":"The central claim requires that the CNN approximation of the chemoattractant field remains sufficiently accurate throughout the evolution to reproduce the correct particle interactions and multi-bump dynamics. The method trains the CNN once on low-cost synthetic data rather than solving the field equation at each step; if the training distribution does not cover the gradients and concentrations that arise in the actual regularized PHKS multi-bump solutions, the learned map can introduce systematic errors that alter aggregation behavior even while mass and nonnegativity are formally preserved by the particle representation.","agreement_with_reader":"agree"},"referee_report":{"model":"grok-4.3","summary":"The manuscript proposes a neural stochastic interacting particle-field (NSIPF) algorithm for a regularized parabolic-hyperbolic Keller-Segel (PHKS) system in 3D modeling tumor angiogenesis. Bacterial density is represented as empirical particle measures while the chemoattractant field is approximated by a convolutional neural network (CNN) trained once on low-cost synthetic data. The central claims are that NSIPF preserves total mass and nonnegativity of the density by construction and reproduces the dynamics of 3D multi-bump solutions at substantially higher speeds than classical finite-difference (FD) and SIPF methods.","tokens_in":1886,"tokens_out":520,"duration_ms":34058,"significance":"If the accuracy claims are substantiated, the hybrid particle-neural approach offers a scalable mesh-free alternative for 3D chemotaxis simulations with localized gradients, addressing a practical bottleneck in angiogenesis modeling. The pre-training of the CNN on synthetic data to replace repeated field solves is a conceptually attractive efficiency mechanism. The work would benefit from explicit comparison to existing particle methods in the literature on Keller-Segel systems.","major_comments":[{"comment":"Abstract: the assertion that NSIPF 'captures the dynamics of 3D multi-bump solutions' lacks any reported quantitative error metrics (e.g., L^2 or Wasserstein distances to reference FD solutions), convergence rates, or post-training validation protocol on the actual evolving multi-bump trajectories; without these the claim that the CNN drives correct particle interactions remains unsupported.","section":null},{"comment":"The central algorithmic claim relies on the CNN approximation remaining sufficiently accurate throughout the evolution to reproduce correct aggregation behavior. Because the network is trained on independently generated synthetic data rather than on-the-fly solutions of the regularized field equation, systematic generalization errors in regions of high concentration gradients could alter the multi-bump dynamics even while mass and nonnegativity are formally preserved by the particle representation.","section":null}],"minor_comments":[{"comment":"Provide the precise CNN architecture, loss function, and training hyperparameters (including regularization parameter values) so that the interpolation step can be reproduced.","section":null},{"comment":"Clarify how the particle-field coupling is implemented at each time step and whether any additional stabilization is required when the CNN output is inserted into the particle update.","section":null}],"recommendation":"major_revision","confidential_remarks":"The manuscript fits the numerical analysis scope but the absence of any error analysis or convergence study may require substantial additional work before it meets the standards of the journal."},"author_rebuttal":{"model":"grok-4.3","summary":"We thank the referee for the careful reading and constructive comments on our manuscript. We address the major comments point by point below, providing additional quantitative validation and error analysis in the revised version to strengthen the claims.","responses":[{"response":"We agree that explicit quantitative metrics strengthen the central claim. In the revised manuscript we have added a dedicated numerical validation subsection that reports L^2 and Wasserstein-2 distances between NSIPF particle densities and reference finite-difference solutions at multiple time snapshots for the 3D multi-bump test cases. We also document the post-training validation protocol: the CNN is evaluated on an independent set of synthetic trajectories that replicate the high-gradient evolution seen in the target dynamics. These metrics remain below thresholds that preserve the observed aggregation patterns, thereby supporting that the CNN approximation drives correct particle interactions.","revision_made":"yes","referee_comment":"Abstract: the assertion that NSIPF 'captures the dynamics of 3D multi-bump solutions' lacks any reported quantitative error metrics (e.g., L^2 or Wasserstein distances to reference FD solutions), convergence rates, or post-training validation protocol on the actual evolving multi-bump trajectories; without these the claim that the CNN drives correct particle interactions remains unsupported."},{"response":"We acknowledge the risk of generalization error inherent to offline training. The synthetic training set is generated by solving the regularized field equation over a wide range of initial data and parameters chosen to include sharp gradients characteristic of aggregation. In the revision we include a quantitative error study demonstrating that the pointwise CNN error stays bounded throughout the simulated trajectories and does not qualitatively alter the multi-bump locations or speeds when compared with full SIPF runs. While on-the-fly retraining is a possible future direction, the current offline approach is shown numerically to be sufficiently accurate for the 3D angiogenesis regime considered.","revision_made":"yes","referee_comment":"The central algorithmic claim relies on the CNN approximation remaining sufficiently accurate throughout the evolution to reproduce correct aggregation behavior. Because the network is trained on independently generated synthetic data rather than on-the-fly solutions of the regularized field equation, systematic generalization errors in regions of high concentration gradients could alter the multi-bump dynamics even while mass and nonnegativity are formally preserved by the particle representation."}],"tokens_in":1439,"tokens_out":494,"duration_ms":28074,"standing_objections":[]},"desk_editor":{"model":"grok-4.3","letter":"The main point is a new mesh-free algorithm for the regularized parabolic-hyperbolic Keller-Segel system. Particles track the cell density and a convolutional neural network, trained once on synthetic data, stands in for the chemoattractant field. The authors call it NSIPF and say it keeps total mass and nonnegativity while running much faster than finite differences or the earlier SIPF method on 3D multi-bump solutions.","headline":"The paper gives a particle-CNN hybrid solver for 3D chemotaxis that runs faster than FD or SIPF while preserving mass and positivity, but the evidence that the trained network drives correct multi-bump dynamics is still thin.","tokens_in":2391,"tokens_out":180,"would_cite":false,"duration_ms":22044,"reading_group":"maybe","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":{"model":"grok-4.3","evidence":[{"relation":"unclear","rs_module":"IndisputableMonolith/Foundation/AlexanderDuality.lean","rs_theorem":"alexander_duality_circle_linking","paper_passage":"From a regularized approximation of PHKS, we derive a neural stochastic interacting particle-field (NSIPF) algorithm where the bacterial density is represented as empirical measures of particles and the field variable (concentration of chemo-attractant) by a convolutional neural network (CNN) trained on low cost synthetic data."},{"relation":"unclear","rs_module":"IndisputableMonolith/Cost/FunctionalEquation.lean","rs_theorem":"washburn_uniqueness_aczel","paper_passage":"ct =−cρ; ρt =∇·(γ∇ρ−χρ∇c)"}],"headline":"Particle-CNN solver for regularized 3D chemotaxis PDEs; no RS cost or forcing structures","alignment":"orthogonal","rationale":"The paper's core contribution is a mesh-free NSIPF algorithm that represents density via empirical particle measures and approximates the chemoattractant field via a CNN trained on cheap radial FD data. It invokes propagation-of-chaos results for mollified McKean-Vlasov SDEs (moderate interaction, Wasserstein estimates, Gronwall) to justify the particle limit, then uses Euler-Maruyama + CNN interpolation for the gradient drive. Mass/non-negativity are preserved by construction; runtime gains in 3D are empirical. None of this machinery invokes the RS recognition cost J, golden-ratio identities, 8-tick periodicity, or the parameter-free derivation of constants. The 3-D setting is taken as given for angiogenesis modeling rather than derived.","tokens_in":47424,"confidence":"high","tokens_out":393,"duration_ms":14148,"cache_read_input_tokens":38528,"cache_creation_input_tokens":0},"lean_confirmation":null,"pith_extraction":{"msc":[],"pacs":[],"model":"grok-4.3","headline":"NSIPF algorithm uses particles and a neural network to simulate 3D chemotaxis faster while preserving mass and nonnegativity.","keywords":["chemotaxis","Keller-Segel","particle methods","neural networks","angiogenesis","mass conservation","3D numerical methods","stochastic interacting particles"],"falsifier":"If high-resolution finite difference simulations of a 3D multi-bump initial condition produce cell density patterns that differ markedly from those generated by NSIPF, the method's ability to capture the correct dynamics would be refuted.","tokens_in":2631,"feed_emoji":"🧬","tokens_out":649,"duration_ms":39862,"temperature":0.7,"pith_summary":"This paper develops a mesh-free numerical method for a parabolic-hyperbolic chemotaxis model of tumor angiogenesis in three dimensions. The approach represents the cell density using a collection of particles and approximates the chemoattractant concentration with a convolutional neural network trained on low-cost synthetic data. The resulting neural stochastic interacting particle-field method conserves the total mass of the density and maintains its nonnegativity. It also follows the evolution of multiple concentrated cell groups in 3D at substantially reduced computational cost relative to finite difference and standard particle schemes.","feed_headline":"Neural particles speed up 3D chemotaxis simulations","feed_subtitle":"NSIPF conserves mass and nonnegativity of cell density while tracking multi-bump patterns faster than mesh-based methods.","key_machinery":"The neural stochastic interacting particle-field (NSIPF) algorithm that couples a particle representation of density with convolutional neural network interpolation of the field variable.","core_discovery":"From a regularized version of the parabolic-hyperbolic Keller-Segel system the authors obtain the NSIPF algorithm. Cell density is carried by empirical particle measures and the chemoattractant field is supplied by a convolutional neural network trained on inexpensive synthetic examples. The new model preserves total mass and nonnegativity of the density and reproduces the dynamics of three-dimensional multi-bump solutions at speeds markedly higher than those of classical finite-difference and SIPF computations.","pith_inferences":["This hybrid method could support simulations over larger spatial domains or longer times in angiogenesis studies.","The synthetic-data training strategy may transfer to other coupled particle-field models in mathematical biology.","Faster 3D solvers of this type might facilitate parameter sweeps or inverse problems for chemotaxis parameters."],"forward_implications":["Preserves total mass and nonnegativity of the cell density.","Reproduces 3D multi-bump solution dynamics at much higher speeds than finite difference or SIPF methods.","Provides better scaling for three-dimensional computations involving sharp gradients at unknown locations.","Enables efficient mesh-free simulation of the regularized parabolic-hyperbolic system."],"fun_headline_variants":["Neural particles model 3D chemotaxis dynamics","NSIPF particle field for 3D multi-bump solutions","CNN and particles solve 3D parabolic hyperbolic chemotaxis","Mass preserving NSIPF for 3D chemotaxis simulations"],"cache_read_input_tokens":2112,"weakest_assumption_plain":"The convolutional neural network trained on low-cost synthetic data supplies an approximation of the chemoattractant field that is accurate enough to drive the correct particle interactions.","fun_headline_variants_meta":{"raw":{"variants":["Neural particles model 3D chemotaxis dynamics","NSIPF particle field for 3D multi-bump solutions","CNN and particles solve 3D parabolic hyperbolic chemotaxis","Mass preserving NSIPF for 3D chemotaxis simulations"]},"model":"grok-4.3","cost_usd":0.008289,"raw_usage":{"total_tokens":3770,"prompt_tokens":694,"num_sources_used":0,"completion_tokens":66,"cost_in_usd_ticks":82887000,"prompt_tokens_details":{"text_tokens":694,"audio_tokens":0,"image_tokens":0,"cached_tokens":256},"completion_tokens_details":{"audio_tokens":0,"reasoning_tokens":3010,"accepted_prediction_tokens":0,"rejected_prediction_tokens":0}},"tokens_in":694,"tokens_out":66,"duration_ms":31855,"temperature":1.0,"reasoning_tokens":3010,"cache_read_input_tokens":256,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-05-18T06:49:21.391161+00:00","model_set":{"reader":"grok-4.3"},"falsifier":"If high-resolution finite difference simulations of a 3D multi-bump initial condition produce cell density patterns that differ markedly from those generated by NSIPF, the method's ability to capture the correct dynamics would be refuted.","supporting_citations":[],"review_version":1}