{"id":"3f3e1554-57fe-4233-b25d-26bfdf737110","arxiv_id":"2506.12803","paper_version":2,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":4,"one_line_summary":"Slime molds confined by blue light explore in many directions but escape by settling into the single contraction mode that most efficiently pumps mass along the longest axis of the trap.","lead":"Researchers trapped slime molds in blue-light shapes and watched how they escaped. The molds first poked out in many directions, then settled into one internal pumping pattern that pushed mass out along the shape's longest axis.","discovery_kind":"new_application","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Post-hoc insertion of the observed hole is what aligns the theoretical 'most efficient' contraction mode with the observed escape; without it the default model predicts the wrong axis. The efficiency-selection claim is therefore not independently supported.","rationale":"The paper has real strengths: a clean behavioral dataset across 32 plasmodia and five shapes, a plausible mechanical narrative connecting peristaltic wave orientation to protrusion growth, and reproducible data/code promised in the repository. The concern is not with the behavioral observation but with the mechanistic conclusion drawn from it. The default theoretical network does not reproduce the observed escape direction; agreement is achieved by inserting the observed hole, a feature that is itself part of the dynamics being explained. That makes the efficiency calculation calibrative, not predictive. The single-exemplar PCA amplifies the problem: without replication we cannot distinguish a general principle of mode selection from a property of one chosen video. The reader's weakest-assumption analysis identifies the same post-hoc, topology-dependent mechanism, so I agree with the reader. The proposed cross-validation—predicting escape from pre-escape network topology across shapes and specimens—would settle whether the post-hoc objection is fatal. Because the concern is concrete, testable, and confined to the interpretive layer, it supports the existing CONDITIONAL verdict rather than moving to rejection.","tokens_in":15489,"tokens_out":7239,"duration_ms":90935,"concrete_test":"Segment each recorded plasmodium's network from the high-resolution or 2-min data at a time before the escape protrusion extends beyond 1 mm, and run the SVD dissipation-efficiency model on that pre-escape network for at least three specimens per trap shape. Compare the predicted top transport axis with the observed escape angle; repeat for the hexagon without manually inserting the hole. If agreement requires using the escape-time hole in the majority of cases, or predicted angles deviate from observed escapes by more than ~20°, the post-hoc objection stands and the efficiency-selection claim is not established.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The central claim that P. polycephalum settles on the contraction mode most efficient for transport is the least secure part of the paper. The only direct test is the hexagon in Fig. 4, and it succeeds only after the observed hole is inserted into an otherwise uniform relative-neighborhood graph. With default geometry the top theoretical mode is vertical (Fig. 4a), while the observed escape is roughly horizontal (160°); the text says these are consistent, but the experimental PCA mode 1 is horizontal and PCA modes 2–3 are vertical (Fig. 3c), so the default comparison is already mismatched. After adding the hole, the dominant mode rotates to 120° and the second to 30°, and the paper appeals to their combination to reach ~160°—an approximate, post-hoc fit rather than a prediction. The hole is part of the same organism's trajectory and may itself result from the exploration dynamics, so using it as a fixed input conflates cause and effect. In addition, the PCA-mode analysis is shown for a single hexagon exemplar, with no statistics connecting mode 1 to escape. Thus, while escape along the longest axis is well supported by 32 plasmodia, the stronger efficiency-optimization claim rests on one calibrated example.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper studies how the slime mold Physarum polycephalum explores and escapes from blue-light traps of different polygonal shapes. The authors report three main findings: (1) across 32 plasmodia in five trap shapes, escape consistently occurs along the longest axis of the shape; (2) during a ~1.5 h exploration phase, protrusions emerge around nearly the entire boundary and their angular positions correlate qualitatively with the orientation of the peristaltic contraction wave; and (3) principal component analysis of contraction patterns shows switching among contraction modes, and a theoretical dissipation-based network model is claimed to show that the organism ultimately settles on the contraction mode most efficient for fluid transport, which coincides with the escape direction. The theoretical identification is obtained by computing SVD modes of a dissipation-ratio efficiency matrix on a uniform relative-neighborhood graph of the trap, and then inserting the observed hole in the hexagonal network to rotate the predicted dominant mode toward the observed escape angle.","tokens_in":15769,"tokens_out":4741,"duration_ms":58070,"significance":"If the central mechanistic claim were fully supported, the paper would be an important advance: it connects decision-making in a non-neural organism to measurable mechanical quantities (peristaltic flow, contraction modes, dissipation) and makes a falsifiable prediction about how network topology should reshape transport modes. The escape-along-longest-axis result is well supported by a large sample (32 plasmodia, five shapes) and is a clean, quantitative behavioral finding. The paper also promises to release data and code. However, the stronger claim—that the organism settles on the transport-optimal contraction mode—rests on a single hexagonal exemplar, on qualitative visual comparisons between experimental and theoretical modes, and on a theoretical model that is adjusted post hoc by inserting an observed hole. These limitations are substantial enough that the paper cannot be accepted without additional evidence or a substantial softening of the claim.","major_comments":[{"comment":"The central claim that the organism settles on the most efficient transport mode is not independently supported. The default relative-neighborhood graph of the hexagonal trap predicts the first transport mode along the vertical axis, while the observed escape is approximately 160° (roughly horizontal); the text states that the default orientations are \"consistent\" with the experimental PCA modes, but Fig. 3c shows the dominant experimental mode 1 as horizontal and modes 2–3 as vertical, so the ordering is actually mismatched. Agreement is achieved only after manually inserting the observed large hole into the network, which rotates the first mode to 120° and the second to 30°, and the text then appeals to their combination to reach ~160°. This is a post hoc calibration, not a prediction, and no quantitative criterion is given for how the two modes combine to yield the escape angle. Because the hole arises from the same exploratory process under study, using it as a fixed input conflates cause and consequence.","section":"Confinement sets contraction mode optimized for transport (Fig. 4a–b)"},{"comment":"The PCA-based mode-switching analysis and the association of mode 1 with escape are presented for a single plasmodium in a hexagonal trap. No replicate numbers, error bars, or statistical tests are given for the mode structures, the temporal activity traces, or the mode-versus-protrusion comparison across the 32 recorded plasmodia or across the five shapes. The statement that \"the transition to escape behavior is primarily associated with mode 1\" is therefore an observation about one individual, not a general result. The paper should either provide multi-individual PCA analyses (at least two or three per shape) or explicitly reframe the PCA section as a case study that motivates, rather than establishes, the optimization claim.","section":"Varying activation of contraction modes during exploration (Fig. 3)"},{"comment":"The theoretical efficiency metric depends on assumptions that are not tested or justified for Physarum. The model assumes a uniform relative-neighborhood graph with equal edge lengths and radii, and defines efficiency as the ratio of Hagen–Poiseuille dissipation to oscillatory-flow dissipation. This ratio is introduced as an axiom rather than derived from measured energy expenditure, and the predicted \"most efficient\" mode is likely sensitive to this choice. No sensitivity analysis is provided for the grid spacing, the graph construction rule, the cost function, or the size and placement of the inserted hole. Given that the key theoretical result is obtained only after modifying the default network, the paper should demonstrate that the predicted dominant mode is robust to reasonable perturbations of these modeling choices.","section":"Transport efficient flow modes (Methods)"},{"comment":"The claim that protrusion growth aligns with peristaltic wave orientation is supported only by qualitative visual comparison of a single hexagonal example and single examples for the other shapes in the supplementary figures. No quantitative correlation, circular statistics, or hypothesis test is reported between the angular position of protrusions and the wave orientation over time. The assertion that the wave \"stabilizes\" around 150° before escape is based on inspection of one time series. This is a load-bearing part of the argument that peristaltic waves drive exploration, so a quantitative measure (e.g., circular correlation, or a comparison of wave-orientation distributions during successful versus retracted protrusion events) is needed.","section":"Orientation of peristaltic wave aligns with location of protrusion (Fig. 2)"}],"minor_comments":[{"comment":"The word \"polygones\" should be \"polygons.\"","section":"Abstract"},{"comment":"\"principle contraction modes\" should be \"principal contraction modes\" in several places, including the Fig. 3a caption and the section heading \"Varying activation of contraction modes during exploration.\"","section":"Fig. 3a caption and text"},{"comment":"\"Single Value Decomposition\" should be \"Singular Value Decomposition\" (two occurrences).","section":"Transport efficient flow modes and Discussion"},{"comment":"Reference [19] has a stray leading comma: \"[19] , B. Y. Hayden\" should be \"[19] B. Y. Hayden.\"","section":"Bibliography"},{"comment":"The caption mentions \"light orange shaded data points,\" but the figure description does not define the shading in a way that would allow a reader to distinguish exploration from escape protrusions in grayscale printing; consider adding a legend or marker-style distinction.","section":"Fig. 1d caption"}],"recommendation":"major_revision","confidential_remarks":"The paper is from a well-known group and the behavioral finding is solid, but the manuscript overreaches in its central mechanistic claim. The authors may be able to fix this either by providing additional experimental replicates for the PCA and mode-escape association, or by substantially rephrasing the conclusion as a case study and model-illustration rather than a general principle. The reviewer sees no indication of citation manipulation or related concerns; the main issue is the gap between evidence and claim. I would encourage the editor to give the authors the opportunity for a major revision rather than a rejection."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Colleague,\n\nWhat you should know first: this paper has a solid experimental result and a speculative theoretical interpretation. The result is that P. polycephalum, confined by blue light in five different polygonal shapes (32 plasmodia total), escapes along the longest internal axis of the trap, while its exploratory protrusions appear all around the boundary. That is a well-quantified extension of earlier work from the same group, and the statistics look adequate for that claim.\n\nThe genuinely new part is the attempt to connect this behavior to contraction modes. The paper shows, in one detailed hexagon example, that the peristaltic wave direction fluctuates among several axes during the exploration phase and stabilizes in the escape direction about 20 minutes before escape. PCA on the contraction pattern reveals a switch among dominant modes, with the eventual mode aligning with the escape direction. That narrative is plausible and visually supported.\n\nThe soft spot is the theoretical claim that the organism 'settles on the contraction mode most efficient for transport.' The default network model, a uniform relative-neighborhood graph, yields a vertical most-efficient mode, while the observed escape is roughly horizontal. The agreement is only restored by inserting the experimentally observed hole into the network; then the dominant theoretical mode rotates to 120°, and the paper combines the first two modes to get close to 160°. That is post-hoc fitting rather than prediction. The hole is part of the same trajectory and may itself be a consequence of the exploration dynamics, so using it as a fixed input weakens the causal story. The efficiency measure is also a ratio of two dissipation terms from a prior model, not a directly measured quantity.\n\nThe experimental escape result stands on its own. The efficiency-optimization narrative should be presented as a hypothesis that needs independent testing, not as the conclusion. The paper would be much stronger with PCA analysis on multiple replicates, a sensitivity analysis of the theoretical modes to network topology, and ideally a prediction on a new shape made before the topology is used.\n\nThis paper deserves serious peer review. The experimental contribution is real, and the theoretical framework is worth discussing. I would send it to review with a request for major revision, focusing on the single-exemplar PCA and the post-hoc topology match.","headline":"Escape along longest axis is solid; efficiency-optimization claim is post-hoc.","tokens_in":16262,"tokens_out":3329,"would_cite":true,"duration_ms":36558,"reading_group":"maybe","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":[],"pacs":[],"model":"deepseek-v4-flash","headline":"A trapped slime mold's escape route is set by the contraction mode that moves fluid most efficiently.","keywords":["Physarum polycephalum","blue light confinement","peristaltic contraction waves","decision-making without a brain","transport efficiency","dissipation modes","principal component analysis","escape behavior"],"falsifier":"Take a trap whose longest axis is not the mode the dissipation model ranks most efficient, and watch which route the plasmodium takes: if it escapes along the longest axis anyway, the claim that confinement selects the transport-optimal mode is falsified. Alternatively, compute the model's top efficiency mode from the actual network topology before escape and check whether the dominant PCA contraction mode at the moment of escape matches it; a systematic mismatch would settle against the claim.","tokens_in":15288,"feed_emoji":"🦠","tokens_out":13591,"duration_ms":147325,"temperature":0.7,"pith_summary":"This paper claims that the slime mold Physarum polycephalum makes its escape decision mechanically, not by centralized computation. Confined by blue light inside a polygon, the organism first extends small exploratory protrusions all around the boundary while its peristaltic contraction waves switch among several global modes; only after about 1.5 hours does it lock onto the contraction mode that is most efficient for moving fluid through its network, and that mode sets the escape direction along the longest axis of the trap. The authors connect the observed behavior to a flow-network model in which transport efficiency is the ratio of Hagen-Poiseuille dissipation to oscillatory-flow dissipation, and they show that the top principal-component contraction mode at escape matches the top efficiency mode of that model once the experimentally observed network hole is included. If correct, this provides a concrete mechanical account of decision-making in an organism without a brain, with choices emerging from self-organized flows under environmental constraint.","feed_headline":"Trapped slime mold escapes via its most efficient flow mode","feed_subtitle":"Confined by blue light, it samples many flow patterns before locking onto the one that moves mass best","key_machinery":"The central object is the transport-efficiency spectrum of the flow network. For each tube the dissipation is $D_{ij}=Q_{ij}^{2}/\\kappa_{ij}+\\frac{1}{12}\\kappa_{ij}(\\partial V_{ij}/\\partial t)^{2}$, where $Q_{ij}$ is the volume flow rate and $\\kappa_{ij}=\\pi a_{ij}^{4}/(8\\mu l)$ the hydraulic conductance; the first term is the cost of net forward transport and the second the cost of oscillatory flow that does not contribute to net displacement. Using the network relation $\\vec{Q}=\\Gamma\\,\\partial\\vec{V}/\\partial t$ (from the flow-mode framework the paper builds on), the total dissipation becomes a quadratic form in the volume-change rates, and the ratio of the two dissipation terms defines an efficiency measure. Singular value decomposition of this ratio yields orthogonal volume-change patterns, whose associated pressure modes are ranked by transport efficiency. On the experimental side, PCA of detrended pixel intensities decomposes the organism's contractions into standing-wave eigenmodes with time-varying relative amplitudes. The argument is carried by matching the dominant PCA mode at escape to the top SVD efficiency mode, with the relative-neighborhood graph (a planar grid graph with no imposed hierarchy) of the trap, including the experimentally observed hole, being the network topology on which the theoretical modes are computed.","core_discovery":"On the paper's own terms, the discovery is that decision-making in P. polycephalum is an active mechanical process carried by fluid transport. In blue-light traps shaped as hexagons, squares, triangles, zigzag polygons, and stars, plasmodia nearly fill the trap and then spend roughly 1.5 hours making small, short-lived protrusions in many directions; the protrusion that finally escapes is consistently located near the longest internal axis. Phase analysis of bright-field images shows that the peristaltic wave reorients during exploration and stabilizes along the escape axis about 20 minutes before escape. PCA of the contraction patterns reveals distinct standing-wave modes whose relative activities switch over time, with the mode aligned with the escape direction dominating at the transition to escape. A theoretical flow network built on the trap geometry, with dissipation split into a net-transport term and an oscillatory-flow term, yields SVD-ranked pressure modes; the most efficient mode aligns with the observed escape direction once the experimentally observed hole in the network is inserted into the model. The paper concludes that harsh confinement forces the organism to settle into the transport-optimal contraction mode, and that the preceding exploration is an out-of-equilibrium sampling of less efficient modes.","pith_inferences":["The paper does not say this, but the same SVD efficiency spectrum could be used as a forward predictor: in a new trap geometry whose network topology, including holes, is known, the model should predict the escape direction before the organism escapes.","The paper does not say this, but if the dissipation-ratio efficiency truly governs mode selection, then artificially creating or removing a hole in the network should rotate the eventual escape direction in the direction the model predicts.","The paper does not say this, but the observed switching among PCA modes could be interpreted as a stochastic search over the dissipation landscape; recording the switching statistics would let one test whether the organism samples modes in proportion to their predicted efficiency."],"forward_implications":["Across five trap shapes, the escaping protrusion consistently leaves near the longest internal axis, so the longest axis is a behavioral signature of the transport-optimization process.","The peristaltic wave is not locked to one direction during exploration; it switches among long-axis directions and only stabilizes about 20 minutes before escape, so exploratory protrusions track the instantaneous wave direction.","Distinct PCA contraction modes support distinct behavioral phases: modes 2–5 dominate the early exploration while mode 1, aligned with the escape route, dominates the transition to escape.","Network topology, not just the outer confinement shape, sets the efficient transport mode: inserting the observed hole rotates the theoretically dominant mode onto the observed escape angle.","Harsh confinement is what triggers the optimal behavior; freely migrating plasmodia of comparable size move fast enough that a less efficient route would suffice, so optimal transport is a response to constraint, not a default state."],"supporting_citations":[{"why":"It supplies the fundamental network relation between volumetric flow and tube volume changes that the dissipation-mode model is built on.","marker":"[54]"},{"why":"It supplies the per-tube dissipation expression whose Hagen-Poiseuille and oscillatory terms define the transport efficiency ratio.","marker":"[33]"},{"why":"It supplies the PCA decomposition of Physarum contractions into standing-wave eigenmodes with time-varying activities.","marker":"[15]"},{"why":"It establishes that peristaltic waves in Physarum align with the longest network axis, the prior result this escape behavior builds on.","marker":"[2]"},{"why":"It provides the phase-extraction and network-topology-response methods used to measure wave direction and interpret how topology reshapes transport modes.","marker":"[10]"},{"why":"It provides the roughly 120-second contraction period used to average peristaltic wave direction over one cycle.","marker":"[23]"},{"why":"It provides the free-migration speed used to argue that unconfined plasmodia can tolerate less efficient routes, supporting the confinement-triggers-optimality claim.","marker":"[27]"}],"fun_headline_variants":["Slime mold's decision is a mechanical flow choice","Slime mold samples flow modes before picking escape route","Mechanical decision: slime mold switches flow modes to escape","Slime mold's escape is a flow-mode optimization","Slime mold decides by shifting its internal flow patterns"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The load-bearing premise is that the organism's network can be treated as a simple uniform network of tubes, that the fluid-dissipation ratio in the model is the actual cost that makes one contraction mode win, and that adding the experimentally observed hole to that idealized network fairly reproduces the escape direction.","fun_headline_variants_meta":{"raw":{"variants":["Slime mold's decision is a mechanical flow choice","Slime mold samples flow modes before picking escape route","Mechanical decision: slime mold switches flow modes to escape","Slime mold's escape is a flow-mode optimization","Slime mold decides by shifting its internal flow patterns"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000411,"raw_usage":{"total_tokens":2192,"prompt_tokens":1072,"completion_tokens":1120,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":688,"completion_tokens_details":{"reasoning_tokens":1042}},"tokens_in":688,"tokens_out":1120,"duration_ms":10077,"temperature":1.0,"reasoning_tokens":1042,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-07T00:42:13.276735+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Take a trap whose longest axis is not the mode the dissipation model ranks most efficient, and watch which route the plasmodium takes: if it escapes along the longest axis anyway, the claim that confinement selects the transport-optimal mode is falsified. Alternatively, compute the model's top efficiency mode from the actual network topology before escape and check whether the dominant PCA contraction mode at the moment of escape matches it; a systematic mismatch would settle against the claim.","supporting_citations":[{"cited_title":"Wilkinson, M","cited_arxiv_id":null,"evidence_quote":"It supplies the fundamental network relation between volumetric flow and tube volume changes that the dissipation-mode model is built on."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"It supplies the per-tube dissipation expression whose Hagen-Poiseuille and oscillatory terms define the transport efficiency ratio."},{"cited_title":"Fleig, M","cited_arxiv_id":null,"evidence_quote":"It supplies the PCA decomposition of Physarum contractions into standing-wave eigenmodes with time-varying activities."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"It establishes that peristaltic waves in Physarum align with the longest network axis, the prior result this escape behavior builds on."},{"cited_title":"Chen and K","cited_arxiv_id":null,"evidence_quote":"It provides the phase-extraction and network-topology-response methods used to measure wave direction and interpret how topology reshapes transport modes."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"It provides the roughly 120-second contraction period used to average peristaltic wave direction over one cycle."},{"cited_title":"Kuroda, S","cited_arxiv_id":null,"evidence_quote":"It provides the free-migration speed used to argue that unconfined plasmodia can tolerate less efficient routes, supporting the confinement-triggers-optimality claim."}],"review_version":1}