{"id":"7be945f9-6cdb-41f4-bbd0-aa22bfef3ebb","arxiv_id":"2412.02401","paper_version":2,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":3.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":1,"one_line_summary":"Two-loop finite-temperature perturbation theory disagrees with lattice phi^4 data at high temperature, while thermoparticle spectral functions keep the temporal correlator consistent.","lead":"This proceedings paper reports that ordinary finite-temperature perturbation theory fails against lattice simulations of hot phi^4 theory at two loops, and that spectral functions built from 'thermoparticle' excitations match the data. It matters because the standard toolbox for thermal quantum field theory may need a different starting point.","discovery_kind":"review","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The lattice evidence does not isolate real-pole propagators as the cause: Eq. (4) uses a Dyson-resummed self-energy, so the two-loop failure may be an artifact of resummation. The mapping from NRT to perturbation theory is asserted, not demonstrated.","rationale":"The paper's central claim is drastic: any finite-temperature perturbative expansion built from real-pole propagators is inconsistent. The primary quantitative support in this proceedings is the lattice comparison in Sec. 2. My stress-test focuses on whether the comparison actually tests that claim. Equation (4) is exact only if Π is the full 1PI self-energy; the paper then approximates Π at two loops and uses the resummed denominator. This is a non-perturbative reorganization, not a strict second-order expansion. The resulting failure—two-loop worse than one-loop and inverted temperature ordering—could stem from the resummation's sensitivity to the large zero-mode contribution of the small bare mass, rather than from the real-pole analytic structure. The NRT theorem constrains exact scattering states; the paper asserts, without proof, that a perturbative expansion with free-field propagators must respect this constraint. Since the lattice data are from a single ensemble (Ns=16, am0=0.15, g0=1.5) with no error bars and strong coupling, the causal attribution to real poles is underdetermined. A decisive check is to use the strict perturbative expansion of C(z) to O(g0^2) (expanding the denominator in powers of Π) and to repeat at weaker coupling; if the inversion disappears, the claim loses its quantitative support. This does not refute the NRT/Weldon-based argument, but it shows the presented lattice evidence is not conclusive. The reader's conditional verdict already reflects the need for more robust evidence; my concern adds a specific technical reason. Therefore I recommend no change to the verdict.","tokens_in":6045,"tokens_out":10992,"duration_ms":118971,"concrete_test":"Recompute the Fig. 2 comparison with the strict perturbative expansion of C(z) to O(g0^2), i.e., expand the propagator in Eq. (4) as a geometric series in Π and keep only terms through two-loop order, instead of resumming the self-energy in the denominator. If the strict expansion no longer inverts the temperature ordering (or converges), the breakdown identified in Sec. 2 is an artifact of the Dyson resummation and does not support the claim that real-pole propagators are inconsistent. Additionally, repeat at a weaker coupling (e.g., g0=0.1) and larger am0 to check whether the effect persists in a regime where perturbation theory should be controlled.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The central claim (Sec. 4) is that any perturbative expansion using real-pole propagators is inconsistent. The main quantitative support is the lattice comparison in Sec. 2. But the correlator is not computed by a strict perturbative expansion. Eq. (4) expresses C(z) via the propagator denominator p^2 + m0^2 + Π, with Π the 1PI self-energy; the paper then evaluates Π to O(g0^2) and inserts it in the denominator. This is a Dyson-resummed approximation, not a two-loop expansion of C(z); expanding Eq. (4) would produce a geometric series of self-energy insertions, and truncating Π at two loops omits an infinite class of higher-order diagrams while including others. The observed failure (two-loop worse than one-loop, inverted temperature ordering for Nτ=2) may therefore be a property of this resummation scheme, not of the real-pole structure of the free propagators. The NRT theorem constrains exact scattering states; the paper does not prove that a formal perturbative expansion in free fields must satisfy this constraint. The single lattice ensemble (Ns=16, am0=0.15, g0=1.5) has no error bars and strong coupling, so it cannot distinguish a resummation artifact or strong-coupling breakdown from a structural inconsistency.","agreement_with_reader":"partial"},"referee_report":{"model":"deepseek-v4-flash","summary":"The manuscript argues that finite-temperature perturbation theory built on free-field or quasi-particle propagators with purely real poles is inconsistent, invoking the Narnhofer-Requardt-Thirring (NRT) theorem as a non-perturbative obstruction. In lattice phi^4 theory, the authors compare spatial correlator predictions computed from a two-loop self-energy with Monte Carlo data and report that at Ntau=2 the two-loop prediction is worse than the one-loop one and even inverts the temperature ordering. They then extract thermoparticle spectral functions from the spatial correlator and use them to reproduce the temporal correlator, claiming consistency with the lattice data. The paper concludes that perturbative expansions using real-pole propagators are inconsistent and that thermoparticles may provide a consistent starting point for finite-temperature perturbation theory.","tokens_in":6280,"tokens_out":4000,"duration_ms":44491,"significance":"If correct, the central claim would have broad implications for finite-temperature quantum field theory, suggesting that standard perturbative treatments require a fundamentally different starting point. The two-loop lattice perturbation theory computation is a legitimate calculation, and the qualitative deterioration at strong coupling is plausible. The thermoparticle framework is an interesting non-perturbative construction that deserves further quantitative study. However, as presented, the evidence is insufficient to support the strong general claim: it rests on a single lattice ensemble with no error bars, on a Dyson-resummed rather than strict perturbative evaluation, and on an asserted rather than demonstrated mapping from the NRT theorem to formal perturbation theory.","major_comments":[{"comment":"The central numerical evidence is produced by Eq. (4), which evaluates the spatial correlator by inserting a truncated self-energy into the denominator of the propagator. This is a Dyson-resummed approximation, not a strict O(g0^2) expansion of C(z): expanding Eq. (4) would generate a geometric series of self-energy insertions, so the one-loop versus two-loop comparison is a comparison of two different resummation prescriptions. The observed inversion of temperature ordering at Ntau=2 could therefore be a property of this resummation scheme rather than evidence against real-pole free propagators. Please compute the strict perturbative expansion of C(z) to O(g0^2), or justify that Eq. (4) is the correct ordering and demonstrate numerically that the distinction is irrelevant.","section":"Sec. 2, Eq. (4)"},{"comment":"All quantitative claims rest on a single lattice ensemble (Ns=16, am0=0.15, g0=1.5) at strong coupling, with no error bars shown on the Monte Carlo points and no variation of Ns, Ntau, am0, or g0. Without statistical errors one cannot substantiate the statement in Sec. 2 that the deviations are 'statistically significant', and strong-coupling effects or finite-volume artifacts are not separated from the proposed structural breakdown. Please provide error bars and at least one check of Ns dependence and a second parameter set.","section":"Sec. 2, Fig. 2"},{"comment":"The inference from the NRT theorem to the inconsistency of formal perturbation theory is asserted rather than derived. NRT constrains exact scattering states with real dispersion relations, whereas a perturbative expansion is a formal asymptotic construction in terms of free fields; the paper does not prove that such an expansion must satisfy the NRT spectral constraint at each order. The branch-point arguments of Refs. [20-22] are cited, but the logical link between those analytic properties and the lattice mismatch is not demonstrated. Please state precisely the theorem that connects NRT to perturbative expansions and explain how the lattice data tests that theorem rather than merely the convergence properties of lattice perturbation theory at this coupling.","section":"Sec. 1 and Sec. 4"},{"comment":"The thermoparticle spectral functions are fitted to the spatial correlator data from the same lattice ensemble, with parameters that can vary with temperature; the subsequent reproduction of the temporal correlator is therefore a consistency check of a fitted model, not an independent prediction. The manuscript does not report the number of fit parameters, the fit quality, or a comparison with alternative spectral ansatze. Please supply these details and a quantitative assessment of whether the thermoparticle model is identifiable from the data and not simply overfitting.","section":"Sec. 3, Fig. 3"}],"minor_comments":[{"comment":"The notation a^2 Pi(...) in Eq. (4) should be defined consistently; please clarify the lattice units of the self-energy and the renormalization scheme (bare vs. subtractive) used for the two-loop diagrams in Fig. 1.","section":"Sec. 2, Eq. (4)"},{"comment":"The quantity eD_beta(u,s) is introduced without a definition; state explicitly whether it is the Fourier transform of D_beta(x,s) and specify its support and normalization conditions as used in the later fitting.","section":"Sec. 3, Eq. (5)"},{"comment":"In the T -> 0 limit the text writes 'D_m,beta(x) -> 1'; as written this is dimensionally inconsistent. Presumably D_m,beta is normalized so that its integral tends to unity; please clarify.","section":"Sec. 3"},{"comment":"The y-axis label rho_TP(a omega)/a^2 is confusing; please specify the normalization of rho_TP and the units of the plotted quantity.","section":"Fig. 3"},{"comment":"The paper refers to Ref. [24] for all methodological details, but since the present proceedings make claims that go beyond that reference, at least the main definitions, the fitting procedure, and the error analysis should be summarized here for the paper to be self-contained.","section":"General"}],"recommendation":"major_revision","confidential_remarks":"This is a proceedings contribution that advertises the stronger results of Ref. [24]. The main issue is that the proceedings version states a very general claim while presenting only one ensemble and no error analysis, and the Dyson-resummed nature of Eq. (4) is not addressed. If the journal accepts proceedings-style papers, the requested revisions may suffice; otherwise the scope of the claim should be narrowed to match the presented evidence."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"The one thing to know: this is a conference proceedings restating Lowdon and Philipsen's JHEP paper [24]. There is no new derivation, equation, or dataset here. If you have already read [24], this adds nothing. If you have not, it is a clear, honest summary of their argument that finite-temperature perturbation theory built on vacuum-like real-pole propagators is inconsistent, with a lattice phi^4 study as evidence.\n\nWhat it does well: the two-loop lattice perturbation theory calculation is genuine, and the reported qualitative failure at N_tau=2—the two-loop predictions being larger than one-loop and even inverting the temperature ordering—is a striking and thought-provoking result. The NRT theorem and the Bros-Buchholz spectral representation are real constraints, and the paper correctly frames them as non-perturbative obstructions rather than infrared artifacts. It also points explicitly to [24] for details, which is the right move for a proceedings.\n\nWhere it is soft: the stress-test note lands. Equation (4) is not a strict two-loop expansion of the correlator; it is a Dyson-resummed expression with the self-energy evaluated to O(g0^2) in the denominator. Expanding (4) would generate an infinite class of higher-order diagrams, so the observed two-loop failure may well be an artifact of the resummation scheme rather than direct evidence that real-pole free propagators are inconsistent. The mapping from NRT—which constrains exact scattering states—to the legitimacy of formal perturbative expansions in free fields is asserted, not proved. The lattice support is also thin: a single ensemble (Ns=16, am0=0.15, g0=1.5) at strong coupling, with no error bars shown and no finite-volume or renormalization checks. That cannot distinguish a resummation artifact from a structural inconsistency.\n\nThe thermoparticle part deserves a specific caveat: the spectral functions are fitted to the spatial correlator data and then used to \"predict\" the temporal correlator on the same ensemble. That is a consistency check, not a confirmation. It shows the ansatz is flexible enough to describe the data, not that thermoparticles are real or that they dominate.\n\nWho is this for? Someone wanting a quick orientation to the authors' program. It is not a self-contained research paper. If it goes through the normal proceedings peer-review route, it deserves a serious referee—the underlying calculation is legitimate and the central claim is important enough that the ambiguity around resummation versus pole structure should be aired. But I would not cite this proceedings in my own work; cite [24] instead.\n\nRecommendation: send it to review, but the referee should push for the error bars, volume/systematics checks, and a comparison between the Dyson-resummed computation and a strict loop expansion of C(z) before taking the inconsistency claim at face value.","headline":"A well-written proceedings summary of the authors' own JHEP paper, but the lattice evidence does not isolate the real-pole propagator structure as the cause of the breakdown, and the thermoparticle 'prediction' is not an independent test.","tokens_in":6830,"tokens_out":1852,"would_cite":false,"duration_ms":22131,"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":"Finite-temperature perturbation theory built on real-pole propagators is inconsistent.","keywords":["finite-temperature field theory","lattice perturbation theory","phi^4 theory","spatial correlator","NRT theorem","thermoparticles","spectral representation","self-energy"],"falsifier":"Compute the two-loop spatial correlator at $N_\\tau=2$ on a second lattice ensemble with different $N_s$ and bare parameters and check whether the inverted temperature ordering persists; if it disappears or the two-loop prediction recovers the data, the claimed pole-structure obstruction is not the cause.","tokens_in":5811,"feed_emoji":"🌡️","tokens_out":7609,"duration_ms":68608,"temperature":0.7,"pith_summary":"This paper argues that finite-temperature perturbation theory cannot be built on free-field or quasi-particle propagators with purely real poles. It reports lattice $\\phi^4$ results where two-loop spatial correlator predictions deviate more from data at higher temperature, even inverting the temperature ordering at $N_\\tau=2$ relative to one loop. The paper interprets this as a non-perturbative obstruction rooted in the NRT theorem, and shows that thermoparticle spectral functions extracted from spatial correlator data reproduce the temporal correlator where perturbation theory fails. The aim is to establish that a consistent finite-temperature perturbative expansion must start from non-perturbative, thermally broadened degrees of freedom rather than vacuum-like propagators.","feed_headline":"Two-loop heat flips phi^4 perturbation theory","feed_subtitle":"At the hottest lattice, two-loop predictions miss data and invert temperature ordering, pointing to real-pole propagators.","key_machinery":"The mechanism carrying the argument is the NRT theorem, the result that nontrivial scattering states with real dispersion relations do not exist at positive temperature, combined with the finite-temperature generalization of the Källén-Lehmann spectral representation, in which the thermal spectral density contains a discrete delta-component $D_{m,\\beta}(x)\\,\\delta(s-m^2)$ describing thermoparticles. On the lattice, the comparison uses the spatial correlator expressed through the self-energy, where the zero mode of the free-field propagator amplifies the cactus diagram and drives the two-loop failure.","core_discovery":"The central claim is that any perturbative expansion constructed using free field, or quasi-particle propagators with purely real poles, is inconsistent at finite temperature. This is not merely an infrared artefact: in lattice $\\phi^4$ theory at fixed bare parameters, the two-loop spatial correlator predictions deteriorate as temperature increases, and at $N_\\tau=2$ they are worse than the one-loop predictions and even reverse the temperature ordering of the correlators. The deviations trace to the competition between the tadpole and cactus diagrams, which is amplified by the zero mode of the vacuum-like free propagator. The paper further claims that the lattice data are consistent with thermoparticle excitations: spectral functions extracted from the spatial correlator reproduce the temporal correlator at each temperature, unlike two-loop perturbation theory. This supports the view that a consistent perturbative expansion should be formulated in terms of these thermally broadened particle-like excitations.","pith_inferences":["If the NRT-based obstruction is general, then any perturbative calculation at finite temperature that ends with real quasi-particle poles, including screened or resummed expansions, faces the same inconsistency even when the numerics look good.","The thermoparticle framework suggests a practical strategy: extract the thermal spectral density from lattice data in one kinematic setup and feed it into a perturbative expansion, seeding the expansion non-perturbatively.","The single ensemble tested here could be extended to a scan over bare mass and coupling to map where the two-loop breakdown begins, giving a boundary that any proposed remedy must reproduce."],"forward_implications":["Standard finite-temperature perturbative predictions in $\\phi^4$ theory and similar scalar theories that use vacuum-like propagators will fail at sufficiently high temperature, with higher orders making the comparison worse rather than better.","A consistent finite-temperature perturbation theory should replace real-pole propagators with thermally broadened thermoparticle spectral functions, whose parameters must be determined non-perturbatively.","The lattice spatial correlator can serve as a diagnostic: when two-loop predictions invert the temperature ordering, the perturbative expansion has broken down.","Thermoparticle spectral functions extracted from one correlator can predict another, providing a cross-check of the non-perturbative description against independent lattice data."],"supporting_citations":[{"why":"Supplies the NRT theorem, the non-perturbative result that scattering states with real dispersion relations do not exist at positive temperature.","marker":"[18]"},{"why":"Provides the argument that real-pole field propagators undermine the consistency of finite-temperature perturbative expansions.","marker":"[19]"},{"why":"Shows explicitly that finite-temperature perturbation theory with real-dispersion propagators breaks down at a fixed loop order in $\\phi^4$ theory.","marker":"[22]"},{"why":"Is the companion paper containing the full lattice simulation and perturbative comparison, including the thermoparticle spectral extraction.","marker":"[24]"},{"why":"Establishes the general spectral representation and the discrete delta-component form for stable particle contributions that defines thermoparticles.","marker":"[26–29]"},{"why":"Supplies the fitting method, previously applied to lattice QCD correlators, used to extract thermoparticle spectral components and predict the temporal correlator.","marker":"[32, 33]"}],"fun_headline_variants":["Real-pole propagators fail at finite temperature","Two-loop heat flips phi^4 perturbation order","Thermoparticles rescue hot phi^4 lattice data","At N_tau=2, two-loop loses to one-loop"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The argument assumes that the NRT theorem's prohibition on real-dispersion scattering states applies directly to formal perturbative expansions built from free-field propagators, and that the observed lattice mismatch at the single simulated ensemble is caused by that pole structure rather than by the chosen lattice parameters, truncation order, or renormalization scheme.","fun_headline_variants_meta":{"raw":{"variants":["Real-pole propagators fail at finite temperature","Two-loop heat flips phi^4 perturbation order","Thermoparticles rescue hot phi^4 lattice data","At N_tau=2, two-loop loses to one-loop"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000388,"raw_usage":{"total_tokens":1973,"prompt_tokens":800,"completion_tokens":1173,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":416,"completion_tokens_details":{"reasoning_tokens":1107}},"tokens_in":416,"tokens_out":1173,"duration_ms":13470,"temperature":1.0,"reasoning_tokens":1107,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-11T23:29:41.594655+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Compute the two-loop spatial correlator at $N_\\tau=2$ on a second lattice ensemble with different $N_s$ and bare parameters and check whether the inverted temperature ordering persists; if it disappears or the two-loop prediction recovers the data, the claimed pole-structure obstruction is not the cause.","supporting_citations":[{"cited_title":"Narnhofer, M","cited_arxiv_id":null,"evidence_quote":"Supplies the NRT theorem, the non-perturbative result that scattering states with real dispersion relations do not exist at positive temperature."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Provides the argument that real-pole field propagators undermine the consistency of finite-temperature perturbative expansions."}],"review_version":1}