{"id":"fda26c5d-0520-4266-8bb5-862a5cfd8f73","arxiv_id":"2411.16882","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":0,"one_line_summary":"Periodic obstacle arrays lock the polarization of aligning active matter to substrate symmetry directions, and anisotropic arrays produce coupled or uncoupled lane states.","lead":"This paper simulates self-propelled particles with aligning collisions moving through periodic obstacle grids, and shows how the grid symmetry steers the direction of collective motion. It maps out three flow states as the grid becomes more anisotropic: freely oriented flow, coupled lanes, and independent lanes.","discovery_kind":"new_application","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Lane-state phase diagram rests on a single box with exactly 8 obstacle rows; no system-size variation in lane number is reported.","rationale":"The paper's novel contribution is the lane-state phase diagram for anisotropic obstacle arrays. The evidence for the coupled-lane phase is the interface probability plateau at ≈0.2 in Fig. 6(b), but with only 8 lanes and thus 7 adjacent pairs, this plateau corresponds to about 1–2 domain walls. That is exactly the kind of small-number signature that can arise from finite-size fluctuations rather than a thermodynamic phase. The reader's weakest_assumption identifies this same gap: no systematic variation of system size or number of rows is reported. I agree with that assessment. The finite-size concern is load-bearing because the central claim explicitly generalizes to 'a wide range of active matter or flocking models' and to 'periodic substrates' generally, yet the phase diagram is built from a single Np,y=8 geometry. A concrete test with doubled box size and doubled row count, holding obstacle spacing and per-lane density fixed, would settle whether the interface probability and domain-wall count scale as expected for a genuine phase. Other potential concerns—such as the transition threshold being unchanged by obstacles—are empirically supported within the tested range (Fig. 3) and are less central. The paper is a solid numerical study, but the lack of this finite-size check justifies the CONDITIONAL verdict; no verdict change is needed.","tokens_in":10892,"tokens_out":8410,"duration_ms":81825,"concrete_test":"Run the same model with L=100σ and Np,y=16, keeping the obstacle spacing identical to the original runs: for each original (Np,x, Np,y=8) case, use (2·Np,x, Np,y=16) while scaling N by 2 to preserve the nominal per-lane density (e.g., N=1152 for the original N=576). Measure the interface probability and the number of domain walls in the region corresponding to the coupled-lane phase (original Np,x between 15 and 25). If the interface probability remains ≈0.2 and the number of domain walls grows proportionally with the number of lanes, the phase is robust; if the interface probability falls or the number of domain walls stays O(1), the 'coupled lane state' is a finite-size artifact.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The central claim that anisotropic obstacle arrays produce generic coupled and uncoupled lane states is supported only by simulations in a square box of side 50σ with exactly Np,y=8 obstacle rows (Section III.C). The distinction between the coupled-lane phase (interface probability ≈0.2) and the decoupled-lane phase (≈0.5) in Fig. 6(b) is based on only 7 nearest-neighbor lane pairs, so the observed values correspond to roughly 1–2 and 3–4 domain walls, respectively. With so few lanes, the coupled-lane state could be a finite-size remnant of the globally polarized state rather than a distinct phase with a finite density of domain walls. The heat maps in Fig. 8 vary N and Np,x but keep Np,y=8, so they cannot reveal whether the phase boundaries or the interface probability are controlled by the number of lanes. The uncoupled state at Np,x>25 is geometric (free gap smaller than a particle diameter), but the crossover region 15<Np,x<25 and the existence of a well-defined coupled-lane phase in the thermodynamic limit are untested.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The manuscript numerically studies a two-dimensional system of active Brownian particles with short-range steric repulsion and aligning torques, in the presence of periodic square and rectangular obstacle arrays. In the obstacle-free case the authors report a polarization transition near β/D ≈ 25, with a data collapse suggesting βc ∼ μD/(v0f). For square obstacle arrays they find that the polarization transition persists at essentially the same parameters, and that dense arrays lock the polarization direction to the substrate symmetry axes (0°, 90°, 180°, 270°). For anisotropic arrays, obtained by increasing the number of obstacles along x while keeping Np,y = 8 rows, they identify a crossover from a quasi-isotropic globally polarized state to coupled one-dimensional lanes and then to uncoupled lanes, characterized by the mean square velocities, the probability of velocity interfaces between neighboring lanes, transient times, and heat maps in the (N, Np,x) plane.","tokens_in":11113,"tokens_out":4641,"duration_ms":42465,"significance":"If the central claims hold, the paper establishes a new and potentially useful mechanism for controlling the direction and spatial organization of collective active flows: passive obstacle geometry can lock polarization to substrate symmetry directions and, for anisotropic arrays, induce lane states with controlled inter-lane coupling. The manuscript's strengths are the use of multiple independent order parameters (individual and collective polarization histograms, mean square velocity components, interface probabilities, transient times), the large number of independent realizations (up to 360 in Fig. 4), and the inclusion of error bars in the main order-parameter plots. The scaling collapse in Fig. 2 over D and N is a genuine attempt to make the transition estimate quantitative. However, the lane-state phase diagram is established for a single box size with exactly eight obstacle rows, which is a load-bearing gap that must be addressed before the phase diagram can be regarded as generic rather than a finite-size artifact.","major_comments":[{"comment":"The identification of the coupled-lane and decoupled-lane states as distinct phases rests entirely on simulations with exactly Np,y = 8 obstacle rows, so the interface probability in Fig. 6(b) is averaged over only seven nearest-neighbor lane pairs. The reported coupled-lane value of about 0.2 and decoupled value of about 0.5 therefore correspond to roughly one or two and three or four interfaces, respectively, a small integer range in which finite-size effects could easily mimic a systematic crossover. Since the heat maps in Fig. 8 vary N and Np,x while holding Np,y = 8, they cannot establish that the phase boundaries or the interface probability are independent of the number of lanes. I request simulations at larger system sizes and at least Np,y = 12 and 16 (or, alternatively, a scaling argument showing that eight rows are sufficient) before the lane-state phase diagram can be accepted as a generic substrate-controlled phase diagram.","section":"Section III.C, Figs. 5, 6, and 8"},{"comment":"The transient-time measurements in Fig. 7(a) include only lanes whose particle density is at least 80% of N/8, a post hoc cutoff that is not varied or justified. Because the number of dense lanes changes with N and Np,x, this filter can systematically alter the reported τ and the conclusion that higher N reduces jamming; the text also defines the steady state only verbally as 'having a lane collectively move to the left or right.' Please report the number of excluded lanes for each parameter set and show that the trends in τ survive reasonable variations of the density cutoff and of the steady-state criterion.","section":"Section III.C, Fig. 7(a)"},{"comment":"The claim that the obstacle array leaves the polarization transition threshold unchanged is supported only by Fig. 3, which shows curves for a single case (N = 240, D = 0.01) without a quantitative estimate of βc or error bars on the transition location. The scaling estimate βc ∼ μD/(v0f) is an order-of-magnitude argument, and the collapse in Fig. 2 is not a finite-size scaling analysis. I ask for a quantitative measure of the transition (e.g., susceptibility peak or Binder cumulant crossing) as a function of obstacle density, or a clear statement that the observed shift is within the resolution of the present measurements.","section":"Section III.A, Eq. (2) and Fig. 3"}],"minor_comments":[{"comment":"The sentence 'the individual particle velocities in Fig. 4(h) exhibit small additional peaks' should refer to Fig. 4(g); Fig. 4(h) is the histogram of the mean polarization direction, not the individual particle velocities.","section":"Section III.B, Fig. 4"},{"comment":"The schematic phase diagram is first introduced as Fig. 5(d), but the text and the Fig. 8 caption refer to it as Fig. 6(d). Please correct the cross-reference.","section":"Section III.C and Fig. 8 caption"},{"comment":"The filling fraction f = Nσ²/L² is used in the scaling estimate without being explicitly defined at first use; please define it when it is introduced.","section":"Section III.A"},{"comment":"The terms 'coupled/decoupled' and 'coupled/uncoupled' lane states are used interchangeably (e.g., 'decoupled lane state' in the text versus 'uncoupled lanes' in Fig. 5 and the abstract); please choose one consistent nomenclature.","section":"Throughout"},{"comment":"Reference [30] displays 'Soft Matter20, 2310' without a space between the journal name and the volume number; please fix the formatting.","section":"References"}],"recommendation":"major_revision","confidential_remarks":"The main risk is that the lane-state phase diagram is based on a single Np,y = 8 system with only seven lane pairs, so the coupled-lane and decoupled-lane classification could be a finite-size effect. I recommend requesting additional simulations with varying Np,y and system size before acceptance; the paper otherwise fits the journal's scope and the central idea is interesting."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Bottom line: the paper reports a genuinely new combination—collective polarization locking and a three-regime lane taxonomy for aligning active matter in periodic anisotropic obstacle arrays. Single-particle directional locking and obstacle-free flocking were known; the collective locking and the coupled-to-uncoupled lane crossover are not in the cited literature. The central claim is plausible and the numerics are mostly careful: multiple independent order parameters, histograms over 360 realizations, error bars, and heat maps over N and Np,x all point the same way. The order-of-magnitude scaling argument for the polarization threshold is rough but honest, and it does not smuggle in the phase boundaries the paper then measures. Self-citations are to prior obstacle-array and aligning-active-matter work and are appropriate.\n\nThe real soft spot, which the stress-test note correctly identifies, is that the entire lane-state phase diagram is built on exactly Np,y=8 obstacle rows in a single box of side 50σ. With only seven nearest-neighbor lane pairs, the quoted interface probabilities of about 0.2 and 0.5 correspond to roughly one to two and three to four domain walls. That is a thin basis for declaring a distinct coupled-lane phase with a finite interface density; it could be a small-system remnant of the globally polarized state. The heat maps vary N and Np,x but keep Np,y fixed, so they cannot resolve this. The transient-time analysis also uses a post hoc 80% lane-density cutoff, which is not justified and could bias the reported relaxation times. No code or data are deposited, so reproducing or extending the simulation results is harder than it should be.\n\nNone of these issues sinks the paper. The polarization-locking result is supported by strong histogram evidence and is likely robust. The lane-state taxonomy is plausible and nicely visualized, but it needs verification against varying Np,y and box size, and the authors should justify or drop the 80% cutoff. I would send this to a serious referee. The referee should ask for those finite-size checks and for code or at least simulation parameters detailed enough to reproduce the main figures.\n\nWho gets value: researchers working on active matter in structured environments, on steering collective motion in microfluidic-like geometries, or on flocking with obstacles. It is a solid subfield contribution, not a paradigm shift.","headline":"Likely real new effect—substrate geometry steering collective polarization in aligning active matter—but the lane-state phase diagram needs finite-size checks before the coupled-lane phase is fully convincing.","tokens_in":11584,"tokens_out":1572,"would_cite":true,"duration_ms":16932,"reading_group":"yes","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":[],"pacs":[],"model":"deepseek-v4-flash","headline":"Periodic obstacle arrays leave the polarization transition threshold unchanged but lock the polarized flow to the lattice symmetry directions, and anisotropic arrays split it into coupled and then independent one-dimensional lanes.","keywords":["active matter","polarization transition","obstacle arrays","directional locking","lane formation","aligning torques","active Brownian particles","flocking"],"falsifier":"Repeat the simulations in larger boxes (e.g., $L = 100\\sigma$ or $200\\sigma$) and with more obstacle rows (16 or 32) at the same filling fraction, and measure the velocity-interface probability and the locations of the quasi-isotropic, coupled, and uncoupled boundaries. If the nearly 0.2 and 0.5 interface probabilities or the phase boundaries shift systematically with system size, the identification of these as generic substrate-controlled phases would be undermined.","tokens_in":10733,"feed_emoji":"🧭","tokens_out":6067,"duration_ms":54774,"temperature":0.7,"pith_summary":"This paper uses numerical simulations of self-propelled particles with steric repulsion and torque-based alignment to ask whether a periodic obstacle substrate changes how a flock forms. It finds that the threshold for the polarization transition is essentially unchanged by obstacles, but the direction of the polarized flow becomes locked to the symmetry axes of a dense square array. When the array is made anisotropic, global two-dimensional polarization gives way first to a coupled lane state with dense and sparse lanes, then to a decoupled state in which each lane polarizes independently left or right. The result matters because it shows that passive geometry can control where active flows go, which could be used to guide collective motion in biological or synthetic active matter.","feed_headline":"Obstacle arrays lock flocking flows to lattice directions","feed_subtitle":"Square posts fix the polarization angle; stretching the array splits the flock into coupled, then independent, lanes.","key_machinery":"The central mechanism is the competition between two time scales: the polarization time $\\tau_P \\sim \\mu L^2/(\\beta N \\sigma^2 v_0)$, the time collisions need to align orientations, and the persistence time $\\tau_D = 1/D$, the time noise decorrelates orientation. The transition occurs when $\\tau_P < \\tau_D$, giving $\\beta_c \\sim \\mu D/(v_0 f)$ with $f = N\\sigma^2/L^2$. The obstacle array does not enter this balance, which explains why the threshold is unchanged; instead it breaks translational symmetry, locking the polarization direction and, for anisotropic arrays, confining motion to channels whose connectivity determines whether lanes couple or act independently.","core_discovery":"In the model studied here, active Brownian particles interacting through short-range repulsion and an aligning torque $\\beta(\\hat{n}_i \\times F_i)\\cdot\\hat{z}$, the transition to a polarized state at critical $\\beta/D$ or $\\beta N$ is not shifted by adding obstacles. Instead, for a dense enough square array the polarization vector and individual velocities become locked to the substrate's $0^\\circ$, $90^\\circ$, $180^\\circ$, and $270^\\circ$ symmetry directions. For rectangular arrays with fixed 8 rows in $y$, increasing the number of columns produces three regimes: a quasi-isotropic globally polarized state, a coupled lane state in which a few dense lanes flow in one direction while sparse lanes coexist and neighboring lanes usually share polarity (interface probability about 0.2), and a decoupled lane state in which particles cannot hop between lanes, each lane flows at uniform density, and the interface probability rises to about 0.5, indicating random per-lane left/right choice.","pith_inferences":["Because the obstacle array only selects the direction of an already-formed polarization rather than changing its onset, the natural design principle that follows is to use the array geometry as a passive switch for collective flow: one could route a flock along desired axes without strengthening the alignment interaction.","The coupled-to-uncoupled transition in the simulations coincides with the obstacle gap becoming smaller than the particle diameter, so the lane-state boundaries should be controlled by this geometric ratio; testing this in other steric models or experiments would turn the phase diagram into a predictive engineering rule.","In the decoupled lane state each lane's left/right choice appears independent, so the 2D system effectively becomes an array of coupled quasi-1D aligning systems; varying the anisotropy tunes the interlane coupling continuously, which should make it possible to study the 1D-to-2D crossover in flocking order in one experimental setup."],"forward_implications":["Below the MIPS threshold, the polarization transition (near $\\beta/D = 25$ or $\\beta N = 70$) is essentially unaffected by adding obstacles, so the substrate does not alter the onset of collective order, only its direction.","For square arrays with $8 \\times 8$ or more obstacles, both individual particle velocities and the system's mean polarization lock to $0^\\circ$, $90^\\circ$, $180^\\circ$, and $270^\\circ$; the locking sharpens as obstacle density rises.","For anisotropic arrays, increasing the $x$-direction obstacle count from about 15 to 25 turns the globally polarized quasi-isotropic state into a coupled lane state with a small number of dense lanes and interface probability near 0.2, and above about 25 obstacles per row into uncoupled lanes with random per-lane left/right polarity and interface probability near 0.5.","In the decoupled lane state, the transient time to reach steady flow decreases with increasing particle number $N$ and angular mobility $\\beta$, because more collisions help lanes escape transient jamming.","A triangular substrate should analogously lock polarization to multiples of $60^\\circ$, as the paper states."],"supporting_citations":[{"why":"Supplies the equations of motion with the aligning torque term and the confined active matter context from which this model is adapted.","marker":"[30]"},{"why":"Provides the prior observation of a polarization transition in interacting polar disks with a different alignment rule, used as the baseline for the scaling-law comparison.","marker":"[41]"},{"why":"Demonstrates single-particle directional locking on periodic arrays, the single-particle analog the collective locking is compared with.","marker":"[7]"},{"why":"Shows directional locking effects for active matter coupled to a periodic substrate, the basis for expecting polarization locking to symmetry directions.","marker":"[16]"},{"why":"Provides another example of microswimmer deflection through obstacle arrays, supporting the generality of locking phenomena.","marker":"[15]"},{"why":"Characterizes the MIPS threshold used to argue that the simulated filling fractions lie well below the regime where motility-induced phase separation would occur.","marker":"[42]"},{"why":"Establishes the broader context of active particles in complex and crowded environments that motivates studying obstacle-coupled active matter.","marker":"[2]"}],"fun_headline_variants":["Square posts pin polarization; stretched arrays split lanes","Obstacle symmetry locks flock; stretching forms lanes","Polarization pins to lattice; stretch splits lanes","Lattice axes lock polarity; stretched arrays make lanes"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The phase diagram and the lane-state interface probabilities are established in a single simulation box of side $50\\sigma$ with exactly eight obstacle rows in the $y$ direction, and the paper reports no systematic variation of box size or row number, so the claim that these constitute generic substrate-controlled phases rests on this small fixed geometry representing an infinite array.","fun_headline_variants_meta":{"raw":{"variants":["Square posts pin polarization; stretched arrays split lanes","Obstacle symmetry locks flock; stretching forms lanes","Polarization pins to lattice; stretch splits lanes","Lattice axes lock polarity; stretched arrays make lanes"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.001249,"raw_usage":{"total_tokens":5109,"prompt_tokens":920,"completion_tokens":4189,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":536,"completion_tokens_details":{"reasoning_tokens":4127}},"tokens_in":536,"tokens_out":4189,"duration_ms":31120,"temperature":1.0,"reasoning_tokens":4127,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-12T12:46:23.150699+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Repeat the simulations in larger boxes (e.g., $L = 100\\sigma$ or $200\\sigma$) and with more obstacle rows (16 or 32) at the same filling fraction, and measure the velocity-interface probability and the locations of the quasi-isotropic, coupled, and uncoupled boundaries. If the nearly 0.2 and 0.5 interface probabilities or the phase boundaries shift systematically with system size, the identification of these as generic substrate-controlled phases would be undermined.","supporting_citations":[{"cited_title":"Canavello, R","cited_arxiv_id":null,"evidence_quote":"Supplies the equations of motion with the aligning torque term and the confined active matter context from which this model is adapted."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Provides the prior observation of a polarization transition in interacting polar disks with a different alignment rule, used as the baseline for the scaling-law comparison."},{"cited_title":"Reichhardt and C","cited_arxiv_id":null,"evidence_quote":"Shows directional locking effects for active matter coupled to a periodic substrate, the basis for expecting polarization locking to symmetry directions."},{"cited_title":"Brun-Cosme-Bruny, A","cited_arxiv_id":null,"evidence_quote":"Provides another example of microswimmer deflection through obstacle arrays, supporting the generality of locking phenomena."},{"cited_title":"Martin-Roca, R","cited_arxiv_id":null,"evidence_quote":"Characterizes the MIPS threshold used to argue that the simulated filling fractions lie well below the regime where motility-induced phase separation would occur."}],"review_version":1}