{"id":"a42f72c9-31d6-4e4f-893b-73038d2f1de1","arxiv_id":"2502.06947","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":5.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":6,"one_line_summary":"A 2+1d simulation of Energetic Causal Sets exhibits a disorder-to-crystal phase transition, mirroring the 1+1d emergence of quasi-particles and showing limit-cycle behavior.","lead":"This Master's thesis extends Energetic Causal Set simulations from 1+1 to 2+1 dimensions and reports a phase transition from disordered dynamics to a crystal-like, time-symmetric phase. A generalist reader might care because it tests whether a quantum gravity program that treats time as fundamentally irreversible can reproduce reversible dynamics in higher dimensions.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Crystal phase may be an artifact of the unprincipled collision-coordinate rule: §6.2 shows the transition appears only after the parent-particle coordinate choice was adopted post hoc.","rationale":"The reader's verdict is CONDITIONAL, and my independent stress-test does not change that verdict: the central claim is plausible but not established. The most load-bearing weakness is not merely the absence of code or error bars, but the paper's own account of how the transition was obtained. Section 6.2 explicitly reports that the parent-particle coordinate rule was the endpoint of a trial-and-error search, that several natural alternatives produced no phase transition, and that the author does not know why this rule works. This is a selection effect inside the algorithm, not a derived consequence of ECS principles. The zero-measure intersection problem in §3.3 makes the finite cross-section necessary, but the finite cross-section alone does not force the parent-particle coordinate rule; that rule is an additional, unexplained modeling choice essential to the reported crystal. Because the central claim depends on this rule, the concern is load-bearing and concrete. The proposed test would settle it by checking robustness across coordinate prescriptions with a quantitative order parameter. I agree with the reader's weakest_assumption, which names the same two choices. No independent support (machine-checked proof, released code, or quantitative diagnostics) offsets the concern, so the paper should remain conditional pending code release, quantitative diagnostics, and a principled justification or robustness test of the collision-coordinate rule.","tokens_in":31813,"tokens_out":3568,"duration_ms":34896,"concrete_test":"Obtain the code and parameter files and run the fully deterministic 2+1d ECS algorithm (σ̄=0.1, dt=σ̄, Tmax=L, L=W=10, 10 families, 10^4 events) under three alternative newborn-coordinate rules: (a) torus-aware midpoint of the two disc centers at first contact, (b) anti-particle parent coordinates, and (c) the point of first tangency between the two discs. Compute a quantitative order parameter (e.g., fraction of events in the dominant family in the final quartile, or a lattice structure factor) over at least 20 independent runs per rule. If a clean 0→1 transition in the order parameter appears only for the parent-particle rule, the reported crystal is a coordinate-rule artifact and the central claim is unsupported; if at least one principled alternative also transitions, the artifact concern is rebutted.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The central claim is that 2+1d ECS exhibits the same disorder-to-order transition as 1+1d, now as a crystal, signaling emergent time-reversible dynamics. For that claim to hold, the transition must be a property of the ECS dynamics, not of the numerical collision prescription chosen to make interactions happen. The paper supplies no such evidence. Because null-ray intersections in 2+1d have zero measure (§3.3, Conclusion), interactions require an ad hoc finite cross-section σ̄. More importantly, §6.2 states that newborn events inherit the parent particle's coordinates rather than the actual collision point, and that this rule was adopted only after 'many, many things' failed. The author explicitly says that using the anti-particle coordinates yields no structure, and the midpoint rule is rejected because it 'constantly add[s] new initial conditions,' preventing the transition. Thus the crystal phase may be manufactured by a coordinate-update rule selected by trial and error, not an emergent ECS phenomenon. The paper's own uncertainty (§5.3: 'we don't know for sure yet'), the absence of code, and the lack of any quantitative order parameter leave this concern unresolved. At minimum, the author must show that a crystal transition is robust to coordinate prescriptions that respect the same conservation laws and torus topology, or derive the parent-coordinate rule from ECS principles.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"This manuscript, a 2022 master's thesis posted to arXiv, extends Cortês–Smolin Energetic Causal Sets (ECS) from 1+1d to 2+1d. Because null-ray intersections in 2+1d form a measure-zero set, the author gives events a finite disc cross-section σ̄, introduces a particle/anti-particle quantum number to conserve the number of families, and implements three parent-selection regimes: fully random, deterministic closest-pasts (a 2d analogue of the 1d rule), and a mixed regime with a tunable randomness probability. The central claims are that (i) the deterministic dynamics show a disorder-to-order phase transition in which a 'crystal' lattice structure emerges, interpreted as the 2+1d analogue of the 1+1d transition to time-reversible dynamics, and (ii) the mixed regime shows limit-cycle-like behaviour, with the proportion of deterministic versus random inputs controlling attraction to and breaking of the ordered phase. The evidence is visual inspection of single simulation runs (Figures 2–13); the causal-network analysis needed to confirm the interpretation is explicitly deferred (§5.3, §8).","tokens_in":32009,"tokens_out":22146,"duration_ms":180012,"significance":"If established, the manuscript would be a useful contribution to the ECS program: it would show that the 1+1d phase transition found by Cortês and Smolin (Refs. 12 and 22) has a higher-dimensional counterpart, supporting the claim that time-reversible dynamics can emerge from fundamentally irreversible laws. The paper earns credit for identifying the zero-measure intersection problem explicitly (§3.3, §8), for checking the random limit against a Poisson expectation with family percentages reported (§4.2), and for reporting specific, checkable observations—notably the five random events whose event numbers and collision times coincide with the breaking of the crystal (§4.3, Figure 11). The author is also candid about what is not known (§5.3, §6.2). The significance is nonetheless prospective: the central observation depends on a collision-coordinate rule adopted post hoc to produce it (§6.2), the evidence is qualitative (single runs, no order parameter, no error bars), the link between the crystal phase and the time-symmetric regime is deferred, and no code is shipped despite an advertised appendix.","major_comments":[{"comment":"The collision-coordinate rule adopted in §6.2 is load-bearing and is acknowledged by the paper to be a post-hoc choice. The new event is assigned the coordinates of the particle parent; the author states that this rule was found only after 'many, many things' failed, that the midpoint rule prevents the transition by 'constantly adding new initial conditions,' and that using the anti-particle coordinates produces no structure—'Why? Again we don't know for sure yet.' Sections 4.1 and 5 additionally state that the 'main objective... was obtaining this phase transition and control it with the parameters.' On the paper's own account, the central observation (the crystal phase transition) was produced by a coordinate prescription selected so that the transition would appear, and the mechanism by which the prescription works is unexplained. Because the rejected alternatives are also plausible (a torus-aware midpoint that respects the conserved momentum data, or an anti-particle rule with a symmetric family-conservation fix), the author must either derive the parent-coordinate rule from the ECS principles of §2.1 or demonstrate robustness of the transition across coordinate prescriptions that preserve the same conservation laws and torus topology; the qualitatively different outcomes of these otherwise-equivalent rules are exactly the test that would separate a dynamical effect from a bookkeeping one.","section":"§6.2; cf. §4.1, §5.1"},{"comment":"The load-bearing claim—that a disorder-to-order phase transition occurs and that its approach is controlled by the proportion of deterministic versus random inputs (Abstract)—is supported only by visual inspection of single runs (Figures 2–13). Section 4.1 concedes that 'these runs are very sensitive to the initial conditions' and that the behaviours 'are to be taken as on average behaviours,' but no averages, ensemble distributions, error bars, or quantitative order parameters are given for the deterministic and mixed runs on which the central claim rests. There is no operational definition of a crystal, of transition time, or of approach speed to the limit cycle; the abstract's statement that the author can 'describe how the proportion of deterministic versus indeterministic inputs... affects the speed of attraction towards the basin of attraction' is not backed by any measurement in §4.3 or §5.2. Moreover, the identification of the crystal phase with the time-symmetric regime rests on a causal-network analysis that the paper explicitly defers (§5.3: 'We don't know for sure yet... What is missing is the causal structure in the set'; §8: 'we have not yet performed an exhaustive analysis of the causal network'). The author should define an order parameter (e.g., the fraction of events in the dominant family, or a lattice structure factor), show its time evolution, and give statistics over many initial-condition draws for each parameter set (number of families, Tmax, σ̄, randomness probability).","section":"§4.1–4.3, §5.1–5.3; Abstract"},{"comment":"The abstract and §8 state that point-like null rays in 2+1d have a zero-measure set of intersections, so 'point particles never interact' and a finite disc cross-section σ̄ is required. The model actually studied is therefore not 2+1d ECS as defined by the four principles of §2.1 but ECS augmented by an interaction prescription, with σ̄ a free parameter whose physical status is not discussed. This matters because the transition is not generic in σ̄: §5.1 and Figure 6 report that for σ̄ ≥ 0.2L no structure forms, and for small σ̄ runs abort for lack of intersections. The author should either justify σ̄ from the ECS principles, provide a stability analysis of the transition across the σ̄ window (using the order parameter of the previous comment), or explicitly restrict the central claim to the extended disc model.","section":"§3.3, §6.1, §8"},{"comment":"The paper is not reproducible in its current form. Section 3 states that 'A version of this code is presented in the Appendix,' but the posted text ends at the References with no appendix, and no repository, DOI, random seeds, or pseudo-random number generator details are given. Since every claim rests on runs of this code, the author should provide the code (or complete pseudocode), the exact parameter values and seeds for each figure, and the event tables underlying Figures 2–13, at least for the deterministic and mixed runs that carry the phase-transition claim.","section":"§3; all of §4–5"}],"minor_comments":[{"comment":"The title 'Quantum Spacetime Leaps' and the abstract's claim of having 'successfully derived the mathematical framework for the 2+1d case' overstate the content: Section 2 states that only the classical version of ECS is treated, and the 'framework' as presented consists of Eqs. (8)–(10) plus the disc-intersection algorithm of Section 3. Consider a title and abstract that match the classical content.","section":"Title and Abstract"},{"comment":"Throughout the text there are typos and garbled phrases that a careful proofread would catch; examples include 'this as nothing to do' (§1.2.3), 'the are two contemplate the dynamics' (§2.1), 'meat' for 'meant' (§3.3), 'poison distribution' for 'Poisson distribution' (§4.2), 'go over σ̄ = 0.5%' (§5.1, presumably σ̄ = 0.5), and 'collisions in the real line' (Abstract, presumably 'in the plane').","section":"Global"},{"comment":"In the caption of Figure 5, plots (c) and (d) are both given Tmax = 20 × L although the text describes four distinct values (10, 50, 100, 200); the 'sandwich effect' invoked there and in §3.1 is never defined. Also, Figure 3(d) shows no crystal within 10,000 events, so the claimed monotonic trend in transition time versus family number rests on a single longer run (Figure 4) with different parameters.","section":"Figure 5 caption; §3.1, §4.1"},{"comment":"The criterion 'within a factor of 2 of the theoretical value' is stated without justification as the test of Poisson consistency; at 100,000 events over 10 families the expected per-family fluctuation is about 1%, so reported values such as family 10 at 6.7% (Figure 7) are several standard deviations away from 10%. A chi-squared goodness-of-fit over the ten families would be a more informative check of the claim that the random code conserves families.","section":"§4.2"},{"comment":"Because interactions are detected on a grid of time step dt = σ̄ (§3.1), the collision time—and hence the coordinates of every newborn event, which are inherited from the parent particle (§6.2)—carry an uncertainty of order σ̄; this quantization should be stated explicitly since it feeds directly into the lattice structure claimed as the crystal phase.","section":"§3.1, §6.2"},{"comment":"The statements that the 3+1d extension is 'fairly straightforward' with no obstacles beyond computational demand are unsupported given the paper's own report that the 2+1d phenomenology is highly sensitive to the randomness prescription (§5.3) and that the interaction prescription had to be re-invented for 2+1d (§6); the zero-measure problem must be re-solved in 3+1d, not merely re-run.","section":"§7, §8"},{"comment":"The dedication, acknowledgements, and the philosophical discussion of Sections 1.1–1.3 are appropriate for a dissertation but should be condensed for an archival physics paper; the excursions into free will, legal responsibility, and societal self-organization are not needed to support the technical claims.","section":"§1.1–1.3, Acknowledgements"}],"recommendation":"major_revision","confidential_remarks":"The manuscript is a master's thesis from 2022, posted as an arXiv paper in 2025, and it reads as one: a long philosophical introduction, dedications, and informal prose are retained. My substantive concern is the evidential gap between the reported phase transition and the presented figures, compounded by the paper's own admission that the collision-coordinate rule was chosen because it produces the transition and that the reason it works is not understood. If the authors can supply the code, a quantitative order parameter with ensemble statistics, and either a derivation or a robustness test of the parent-coordinate rule, the contribution could become solid; in its current form I would not recommend acceptance. There is no indication of citation problems; the ECS literature is properly acknowledged and the relationship to the 1+1d results is described accurately."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Two things to know about this one. First, it is the first serious attempt to carry Energetic Causal Sets into 2+1d, and it gets far enough to reproduce something like the 1+1d disorder-to-order transition, now in the form of a crystal lattice. Second, do not take that crystal at face value. The evidence is a handful of Mathematica plots with no code, no order parameter, no error bars, and the main collision rule was chosen because it produced the transition.\n\nWhat is genuinely good: the thesis is transparent. The author states in Section 5 that the main objective was to obtain the phase transition and control it with parameters, admits that point-like null rays in 2+1d have a zero-measure intersection set, and documents in Section 6.2 that many alternative coordinate rules failed and only the parent-particle rule worked. That honesty is rare and useful. The random-limit test, with roughly even family distributions, suggests the code is not grossly broken, and the algorithm description is concrete enough to reimplement. The mathematics is simple algebra; the citation pattern is appropriate, with all key ECS references in place.\n\nThe soft spots are the load-bearing ones. Section 6.2 is the core: the crystal appears only when newborn events inherit the particle parent's coordinates; the midpoint rule and the anti-particle rule do not produce any structure. The author does not know why, and says so. That is a textbook signature of a modeling artifact. The finite cross-section is also ad hoc, injected to solve the zero-measure problem. It may be the right physical idea, but it needs a justification from ECS principles rather than a trial-and-error success at getting the plot you wanted. The limit-cycle claim is qualitative: no basin statistics, no comparison to DDS theory, just long runs that look periodic.\n\nThis is a master's thesis posted to arXiv, and it reads like one. It is a progress report, not an established result. But it is genuinely new and important to the ECS community, and the author's candor makes it a reasonable starting point. I would send it to a serious referee, and I would make the referee's job concrete: ask for the code, a quantitative order parameter with ensemble statistics, and a robustness check of the crystal across coordinate prescriptions. Without those, the claim remains a conjecture. I would not cite it as evidence for emergent time symmetry in 2+1d yet.","headline":"First serious 2+1d attempt at Energetic Causal Sets, but the crystal phase is likely an artifact of a post hoc collision rule rather than a robust emergent phenomenon.","tokens_in":32660,"tokens_out":3553,"would_cite":false,"duration_ms":32265,"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 2+1d energetic causal set simulation exhibits the same disorder-to-order phase transition observed in the one-dimensional model, with a crystal lattice marking the onset of time-symmetric dynamics.","keywords":["energetic causal sets","quantum gravity","arrow of time","phase transition","crystal lattice","limit cycles","discrete dynamical systems","2+1d simulation"],"falsifier":"Run the deterministic 2+1d algorithm again but assign each newborn event the midpoint of its two parents' centers instead of the particle parent's position, with all other parameters fixed; if a crystal still forms, the reported phase transition cannot depend on the coordinate convention in the way the paper claims.","tokens_in":31472,"feed_emoji":"🧊","tokens_out":6868,"duration_ms":61564,"temperature":0.7,"pith_summary":"The paper proposes a 2+1d computer simulation of Energetic Causal Sets, a model in which time irreversibility and energy-momentum are fundamental and spacetime emerges from the dynamics. Its central goal is to test whether the phase transition seen in the 1+1d model survives in two spatial dimensions: early evolution is disordered and time-asymmetric, and later it settles into an ordered, time-symmetric regime. To overcome the fact that point-like null rays in flat 2+1d almost never intersect, the simulation gives particles a finite radius and defines a newborn event at the particle parent's position. With these conventions the run shows a crystal-like lattice emerging from the disordered phase, and the transition is faster when fewer families or larger cross-sections reduce the random input. The paper also identifies limit-cycle behaviour in the language of discrete dynamical systems, with the ratio of deterministic to random parent selection controlling how quickly the system is captured by the limit cycle.","feed_headline":"2+1d energetic causal sets reach a time-symmetric crystal phase","feed_subtitle":"The same disorder-to-order transition seen in 1+1d appears in two spatial dimensions, suggesting time symmetry is emergent.","key_machinery":"The load-bearing mechanism is the interaction rule and the event-coordinate rule. In flat 2+1d, the set of intersections of point-like null rays has measure zero, so the simulation gives each particle and antiparticle a finite radius (cross-section $\\bar{\\sigma}$) and declares an interaction when two discs overlap; it then defines the spacetime position of the newborn event as the position of the particle parent, chosen consistently so that families are conserved. The parent-selection rule is the 'closest pasts' rule, which compares the causal-past measures of available events and picks the pair with the smallest difference, with a tunable probability of replacing that choice by a random pair. These choices together convert a generically non-interacting point-particle system into one with persistent interactions and a reproducible disorder-to-order transition.","core_discovery":"On its own terms, the paper's discovery is that the 2+1d ECS dynamics reproduces the central feature of the 1+1d model: given purely local, time-irreversible rules for how events create new events, the system nonetheless relaxes into a time-symmetric phase. In two spatial dimensions this phase appears as a crystal, a regular lattice-like arrangement of events in the emergent Minkowski embedding, which the paper identifies with quasi-particle trajectories. The mechanism that makes this possible is treating particles as discs with a finite cross-section and consistently assigning the coordinates of a new event to its particle parent, so that the number of families in the present stays conserved. Once the crystal forms, the system behaves like a discrete dynamical system captured by a limit cycle; the more deterministic the parent-selection rule, the faster the attraction to that cycle. The paper takes this as evidence that reversible dynamics can emerge from fundamentally irreversible laws in higher dimensions, extending the programme beyond the one-dimensional case.","pith_inferences":["An implication the paper leaves implicit: since the zero-measure intersection problem is the stated obstruction, the finite disc radius is doing essential work, and a continuum version of 2+1d ECS would need a different interaction mechanism, such as causal chains meeting through finite regions rather than points.","The particle-parent coordinate rule is the least motivated convention in the model, and the paper reports that a natural alternative, placing newborn events at the midpoint between parent centers, destroys the phase transition; a more symmetric rule that still conserved both particle and antiparticle families would be a testable alternative.","The single-random-event fragility suggests the random input is injected at maximum strength; adopting a gentler source of randomness, for example randomizing the number of boundary crossings before an interaction as in the 1d model, could sharpen the deterministic limit.","If the crystal truly is a limit-cycle basin, the identity of the family that forms the crystal and the time of its formation should be reproducible functions of the initial data across many same-parameter runs, and quantifying that reproducibility would give a sharper test of the claim."],"forward_implications":["The 2+1d model joins the 1+1d model as a place where time-symmetric, effectively reversible dynamics arises at late times from laws that are explicitly time-irreversible at the event level.","The speed of the phase transition is controlled by parameters: fewer initial families, larger particle radius, and smaller Tmax all shorten the disordered phase and hasten the crystal.","Even a single randomly chosen parent event can break an already formed crystal and return the system to the disordered phase, implying that the deterministic limit is approached only as the random probability tends to zero.","Limit cycles, previously identified in 1+1d ECS, also appear in 2+1d; the proportion of deterministic versus random inputs sets the rate of attraction toward the limit cycle.","The crystal phase is accompanied by an apparent loss of discausality, with the total and partial orders aligning so that Minkowski time advances faster per created event."],"supporting_citations":[{"why":"Establishes the 1+1d energetic causal set model and the phase transition to the time-symmetric regime that this paper tries to extend.","marker":"[12]"},{"why":"Provides the general ECS framework with energy-momentum conservation at events, on which the 2+1d simulation is built.","marker":"[11]"},{"why":"Identifies the limit-cycle behaviour of 1+1d ECS in discrete dynamical systems, the baseline that the 2+1d limit-cycle finding is compared with.","marker":"[22]"},{"why":"Supplies the discrete dynamical systems formalism and the claim that limit-cycle attractors are generic.","marker":"[21]"}],"fun_headline_variants":["Time symmetry emerges from irreversible causal set rules","2+1d causal sets show reversible phase from irreversibility","Emergent time symmetry in higher-dimensional causal sets","Irreversible laws yield time-symmetric crystals in 2+1d","Higher-D causal sets reach reversible dynamics from irreversibility"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The crystal phase appears only because particles are given a finite radius and because newborn events are placed at the particle parent's coordinates; change either convention and, by the paper's own report, the phase transition does not occur.","fun_headline_variants_meta":{"raw":{"variants":["Time symmetry emerges from irreversible causal set rules","2+1d causal sets show reversible phase from irreversibility","Emergent time symmetry in higher-dimensional causal sets","Irreversible laws yield time-symmetric crystals in 2+1d","Higher-D causal sets reach reversible dynamics from irreversibility"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.00017,"raw_usage":{"total_tokens":1248,"prompt_tokens":908,"completion_tokens":340,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":524,"completion_tokens_details":{"reasoning_tokens":258}},"tokens_in":524,"tokens_out":340,"duration_ms":3610,"temperature":1.0,"reasoning_tokens":258,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-08T14:14:32.721623+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Run the deterministic 2+1d algorithm again but assign each newborn event the midpoint of its two parents' centers instead of the particle parent's position, with all other parameters fixed; if a crystal still forms, the reported phase transition cannot depend on the coordinate convention in the way the paper claims.","supporting_citations":[{"cited_title":"The Universe as a Process of Unique Events","cited_arxiv_id":"1307.6167","evidence_quote":"Establishes the 1+1d energetic causal set model and the phase transition to the time-symmetric regime that this paper tries to extend."},{"cited_title":"Energetic Causal Sets","cited_arxiv_id":"1308.2206","evidence_quote":"Provides the general ECS framework with energy-momentum conservation at events, on which the 2+1d simulation is built."},{"cited_title":"Reversing the irreversible: from limit cycles to emergent time symmetry","cited_arxiv_id":"1703.09696","evidence_quote":"Identifies the limit-cycle behaviour of 1+1d ECS in discrete dynamical systems, the baseline that the 2+1d limit-cycle finding is compared with."},{"cited_title":"Wuensche, Exploring Discrete Dynamics; The DDLab Manual , Luniver Press, UK, 2011; Attractor Basins of Discrete Networks , Cognitive Science Research Paper 461, Univ","cited_arxiv_id":null,"evidence_quote":"Supplies the discrete dynamical systems formalism and the claim that limit-cycle attractors are generic."}],"review_version":1}