{"id":"646a605a-a0a7-434f-917c-b4e2d64dfab1","arxiv_id":"hep-ph/0403234","paper_version":2,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":0,"one_line_summary":"In a two-loop 2PI simulation of a quark-meson model, the equation of state and a global kinetic temperature reach near-equilibrium values within about one inverse thermal mass, roughly a hundred times faster than full thermal equilibration.","lead":"A numerical simulation of a quark-meson model shows that the ratio of pressure to energy density, and a global kinetic temperature, settle near their equilibrium values within about one inverse thermal mass, long before individual particle modes equilibrate. If the same holds in QCD, it would explain why hydrodynamic descriptions of heavy-ion collisions work so soon after the collision.","discovery_kind":"extension","skeptic_critique":{"model":"selfhosted/deepseek-v4.1-flash","headline":"The load-bearing step is identifying the prethermalized plateau with the equilibrium value: dephasing explains a plateau, not that it sits at w_eq, and the shown runs share one energy density, so the 'to their equilibrium values' claim is not stress-tested.","rationale":"The paper's in-model result is largely supported: a numerical two-loop 2PI solution with no further approximation shows w and T_kin flattening within a few m^-1 while mode quantities take roughly 25-28 m^-1 and full equilibration roughly 95 m^-1 (Figs. 1-4). The dephasing explanation is plausible and previously documented. The reader's identified weak point — the QCD transfer to 'less than 1 fm' — is real but is explicitly framed as an expectation ('We expect such a relation...') and does not bear on the in-model claim; a clearly-labeled extrapolation is not a hidden weakness of the central result. The sharper concern is internal to the in-model argument: the step from 'w becomes time-independent' to 'w reaches its equilibrium value' is the load-bearing one, and the shown evidence does not separate these because all displayed runs share one energy density (Fig. 1 caption; Fig. 4 shows T_kin only for Teq/m = 1.0). Secondary issues the reader notes — the 20% threshold in footnote [24], the single truncation with no order study, coupling independence tested only in h — are real but do not by themselves undermine the separation of scales. The paper's acknowledged limitation (pressure isotropization left for future work, footnote [26]) is stated, not hidden. My concern, if it lands, narrows the abstract's wording rather than overturning the time-scale separation, so the reader's CONDITIONAL verdict is appropriate and I would not move it.","tokens_in":8281,"tokens_out":8717,"duration_ms":157950,"concrete_test":"Add a run C at the same total energy density but with a large initial potential-energy fraction (sigma displaced far from its minimum, small kinetic energy), and repeat at a lower temperature where w_eq deviates strongly from 1/3. Apply the footnote [24] criterion |w(t) - w_eq|/w_eq < 0.2 for t > t_pt. If w(t) reaches within 20% of w_eq by t of order a few m^-1 and stays, the concern does not land and abstract:s0 stands. If it plateaus at a different value and drifts toward w_eq only on t_eq, the 'to their equilibrium values' wording should be softened to 'to a time-independent value.'","verdict_should_be":"UNCHANGED","load_bearing_attack":"The abstract's central claim (abstract:s0) is prethermalization 'to their equilibrium values.' The paper's own mechanism — dephasing of oscillating modes with a dense spectrum — explains why a bulk ratio like w = p/epsilon becomes time-independent quickly; it does not by itself explain why the plateau value equals w_eq. The numerical evidence that the plateau sits near w_eq comes from runs A and B, which by construction share the same energy density and hence the same w_eq, and from the Fig. 3 inset, which varies only h at fixed lambda and fixed energy density. The paper therefore never varies the initial partition of energy between kinetic and potential at fixed total energy density. In a near-conformal regime w_eq is approximately 1/3 and a kinetic-dominated initial state also gives w approximately 1/3, so the observed agreement is easy and may not be diagnostic. If a potential-dominated initial state dephases to a plateau away from w_eq and only drifts there on t_eq, then 'prethermalization to equilibrium values' overstates the result; the defensible statement is prethermalization to a time-independent, approximately thermal value for the initial conditions studied. This is a robustness concern about the central claim, distinct from the reader's QCD-transfer concern.","agreement_with_reader":"partial"},"referee_report":{"model":"selfhosted/deepseek-v4.1-flash","summary":"The manuscript reports a numerical study of far-from-equilibrium quantum field dynamics in a two-flavor linear sigma model with quarks, solved in the two-loop 2PI effective action without further approximation. Three separated time scales are identified: a damping time t_damp ≈ 25–28 m^-1 at which individual momentum-mode occupations lose most of the memory of the initial condition (Fig. 1); a prethermalization time t_pt of order m^-1 at which the bulk ratio w = p/ε and a global kinetic temperature T_kin = E_kin/c_eq become approximately time-independent and close to their thermal values (Figs. 3 and 4); and a thermal equilibration time t_eq ≈ 95 m^-1 at which fermion and scalar mode temperatures coincide (Fig. 2). Chemical equilibration of the fermion and scalar number densities, defined through integrated densities (Eq. 4), is found to proceed on t_ch ≈ t_eq, far slower than prethermalization. The authors argue that an approximately constant equation of state can close a hydrodynamic description long before local thermal equilibrium, and estimate T t_pt ≈ 2–2.5, corresponding to t_pt somewhat below 1 fm for T ≳ 400–500 MeV.","tokens_in":8503,"tokens_out":5885,"duration_ms":102361,"significance":"The result is significant and, if it holds, directly relevant to the early-time description of heavy-ion collisions: it separates the requirement of a closure relation p = p(ε) from the much stronger requirement of local thermal equilibrium, and it does so with a first-principles (2PI) nonequilibrium calculation rather than a kinetic or near-equilibrium approximation. The strengths are concrete: the dynamics is solved numerically without further approximation beyond the two-loop 2PI truncation; the two-stage loss of memory is visible directly in the mode occupations of Fig. 1 and the mode temperatures of Fig. 2; the paper makes quantitative, falsifiable statements (t_damp, t_pt, t_eq, t_ch, and the product T t_pt ≈ 2–2.5) rather than qualitative ones; and the extrapolation to QCD is explicitly labelled as an expectation rather than presented as a derived result. The chemical-equilibration comparison adds a useful, model-internal ordering of time scales.","major_comments":[{"comment":"The abstract claims prethermalization of the equation of state and the kinetic temperature 'to their equilibrium values'. The numerical support for the plateau value, as opposed to the mere existence of a plateau, is thinner than the wording suggests. Runs (A) and (B) of Fig. 1 are constructed to share the same energy density and therefore the same w_eq, and the Fig. 3 inset varies only h at fixed λ and fixed energy density. The manuscript therefore never varies the partition of energy between kinetic and potential (or, equivalently, the zero-mode amplitude) at fixed total energy density, which is precisely the direction in which a dephasing plateau could sit away from w_eq and drift toward it only on t_eq. A single control run with a potential- or zero-mode-dominated initial condition at the same ε, showing that w(t) still reaches the neighbourhood of w_eq on t_pt, would settle this. Absent that, the defensible statement is prethermalization to a time-independent, approximately thermal value for the initial conditions studied, and the abstract should be softened accordingly.","section":"Abstract (abstract:s0), Fig. 3, and footnote [24]","authors_acknowledge":null,"authors_acknowledge_p":null},{"comment":"The kinetic temperature is defined as T_kin(t) = E_kin(t)/c_eq with c_eq = E_kin,eq/T_eq, so T_kin = T_eq at equilibrium by construction; the physical content is the empirical statement that the ratio E_kin/E reaches its thermal value quickly. The text acknowledges this in one sentence, but the abstract's phrasing 'prethermalization of ... the kinetic temperature to their equilibrium values' reads as a dynamical statement about a temperature rather than about an energy ratio. In addition, the manuscript never quotes the numerical value of T_kin(t_pt)/T_eq. The solid curve in Fig. 4 cannot be read to better than the line width, so the reader cannot tell whether the prethermalized value is 1.0, 1.1, or 1.2. Please state the number explicitly and, if it deviates from unity at t_pt, say so rather than leaving it to the figure.","section":"Eq. (3) and Fig. 4","authors_acknowledge":null,"authors_acknowledge_p":null},{"comment":"The momentum cutoff is quoted as Λ/m = 2.86, while Fig. 2 displays mode temperatures out to p = 2.5 m, i.e. within roughly 15% of the cutoff. Since the prethermalization mechanism invoked in the text is dephasing over a dense frequency spectrum, and since the claim that all mode temperatures converge to a common T_eq is read off precisely these high-momentum points, the cutoff sensitivity is load-bearing for the interpretation of Fig. 2. Please state whether the plateau values of w and T_kin, and the value of t_pt, are stable under variation of Λ (for instance Λ/m = 2.0 versus 2.86), or restrict the mode-temperature plot to p ≲ 2 m where the cutoff is not felt.","section":"Footnote [20], Fig. 2, and the dephasing argument","authors_acknowledge":null,"authors_acknowledge_p":null},{"comment":"The estimate T t_pt ≈ 2–2.5 is extracted from a single model with h = λ = 1 and a scale set by the scalar thermal mass, and the transfer to the quark-gluon state is stated as 'We expect such a relation with a similar constant to hold', citing Ref. [25], which is a review by one of the authors rather than a QCD computation. The framing is honest, but the concluding sentence 'For T >~ 400−500 MeV we obtain a very short prethermalization time t_pt of somewhat less than 1 fm' reads as a result of this paper when it is an extrapolation. Please make the status of the constant explicit in that sentence, and state which features of the model (particle content, coupling, the identification of the relevant scale with T) the constant is expected to be insensitive to and which it is not.","section":"Final paragraph and the 'less than 1 fm' estimate","authors_acknowledge":null,"authors_acknowledge_p":null}],"minor_comments":[{"comment":"The two runs (A) and (B) are specified only qualitatively ('high initial particle number density in a narrow momentum range' versus 'closer to a thermal distribution'). Please give the actual initial parameters (zero-mode amplitude, occupation width and normalization, initial time) so that the comparison is reproducible, and state explicitly that both runs have the same energy density.","section":"Fig. 1 caption and initial conditions","authors_acknowledge":null,"authors_acknowledge_p":null},{"comment":"The mode frequency ω_p(t) is defined through the peak of the spectral function. Please state the numerical uncertainty in locating that peak, since the late-time statement that all fermion and scalar mode temperatures agree to a common T_eq is being read off differences of order the symbol size in Fig. 2.","section":"Eq. (2) and Fig. 2","authors_acknowledge":null,"authors_acknowledge_p":null},{"comment":"The argument for a time-independent T_kin rests on total energy conservation ('Since total energy is conserved'). Please report the relative drift of the energy density over the full run as a check on the numerical solution of the 2PI equations, and state how vacuum contributions are subtracted in the energy-momentum tensor.","section":"Energy conservation","authors_acknowledge":null,"authors_acknowledge_p":null},{"comment":"The prethermalization time is defined by |w(t_pt) − w_eq|/w_eq < 0.2 for all t > t_pt. Since t_pt ~ m^-1 is the headline number, please state the criterion in the main text and give the sensitivity of t_pt to the threshold (for example 10% versus 20%).","section":"Footnote [24] and the definition of t_pt","authors_acknowledge":null,"authors_acknowledge_p":null},{"comment":"The statement that the main results are independent of the detailed values of the couplings is supported by the Fig. 3 inset, which varies h only while λ is held fixed. Please say so explicitly, or add the corresponding λ variation.","section":"Coupling independence","authors_acknowledge":null,"authors_acknowledge_p":null},{"comment":"The caption identifies scalars as the upper curves and fermions as the lower curves, but the three (h, T_eq) combinations in the legend are not mapped onto individual curves. Please label the curves directly or add a marker scheme so the reader can associate each pair with its parameters.","section":"Fig. 4 legend","authors_acknowledge":null,"authors_acknowledge_p":null},{"comment":"Ref. [18] is cited for the statement that t_damp depends on the interaction strength, but its stated content is the relation of the partial loss of memory to approximate fixed-point solutions of equal-time evolution equations. Please check that this is the intended citation, or add the reference that actually establishes the coupling dependence of t_damp.","section":"Ref. [18] and t_damp","authors_acknowledge":null,"authors_acknowledge_p":null},{"comment":"There are several typographical and consistency issues to fix: 'looses' for 'loses' (appears twice), inconsistent notation between t_ch and t_ch in the running text, and spacing artifacts in the abstract and author list of the source file.","section":"Typographical","authors_acknowledge":null,"authors_acknowledge_p":null}],"recommendation":"minor_revision"},"author_rebuttal":null,"desk_editor":{"model":"selfhosted/deepseek-v4.1-flash","letter":"Bottom line: the in-model result is solid and worth citing — a clean, quantitative demonstration that bulk quantities settle on a time scale two orders of magnitude shorter than thermal equilibration. The two claims that get quoted most, though, are the weakest supported: that the plateau sits at the equilibrium value, and the 'less than 1 fm' QCD number. What's actually new: they solve the two-loop 2PI equations for a quark-meson model with fermions, no further approximation, and show three well-separated scales — t_pt ~ m^-1 for w = p/eps and T_kin, t_damp ~ 25-28 m^-1 for mode occupancies, t_eq ~ 95 m^-1 for full equilibration. Figs. 1 and 2 make the two-stage loss of memory visible directly. The chemical-temperature comparison in Fig. 4 is a useful addition. They also flag that pressure isotropization is left for future work, which is honest. Where it's soft. First, the stress-test point lands partially: 'prethermalization to their equilibrium values' is not stress-tested. Runs A and B share an energy density, hence the same w_eq, and the Fig. 3 inset varies only h. Footnote 24 defines t_pt by |w - w_eq|/w_eq < 0.2, so 'near w_eq' is built into the criterion rather than independently demonstrated across initial partitions of energy. The figure does show w climbing toward w_eq rather than sitting there, so it isn't trivial, but a potential-dominated start at fixed eps is the test that's missing. Second, the femtometer estimate is explicitly an expectation ('We expect...') resting on the authors' own review [25], not a QCD computation. It's honestly labeled, but it's the headline. Third, no error bars or truncation-order study on any time scale, and 'coupling independence' varies only h, not lambda. These are minor-to-moderate, not fatal. Who it's for: heavy-ion phenomenology and nonequilibrium field theory; anyone who needs a citable statement that hydrodynamization can precede thermalization. Recommendation: send it out. The core time-scale separation is sound and the extrapolations are flagged as such; a referee should press on the initial-condition robustness of the plateau value and on the QCD transfer, not on the in-model result.","headline":"The in-model time-scale separation is solid and citable; the two most-quoted claims — that the plateau sits at the equilibrium value, and the sub-femtometer QCD number — are the least supported.","tokens_in":9202,"tokens_out":2900,"duration_ms":46991},"pith_extraction":{"msc":[],"pacs":["11.10.Wx","12.38.Mh","05.70.Ln"],"model":"selfhosted/deepseek-v4.1-flash","headline":"The paper argues that a far-from-equilibrium quark-meson system fixes its equation of state and kinetic temperature within about one inverse thermal mass, long before full thermal or chemical equilibrium.","keywords":["prethermalization","quark-meson model","2PI effective action","heavy-ion collisions","hydrodynamic equation of state","kinetic temperature","chemical equilibration","nonequilibrium quantum fields"],"falsifier":"A controlled first-principles QCD calculation of an expanding quark-gluon plasma that finds $w=p/$$ or $T$$ still drifting by more than about 20% at $1$$, or that yields a substantially different constant in $T t$$, would undercut the paper's heavy-ion extrapolation; a repeat of the same 2PI calculation with a different truncation or coupling finding no early plateau would undercut the in-model claim.","tokens_in":9576,"feed_emoji":"⚛️","tokens_out":7193,"duration_ms":136086,"temperature":0.7,"pith_summary":"The paper argues that a far-from-equilibrium quantum field theory can acquire a usable equation of state and a global kinetic temperature long before it thermalizes. In a two-flavor quark-meson model evolved with a two-loop 2PI effective action, the ratio $w=p/$$$ and the kinetic temperature $T$$ approach their equilibrium values on a prethermalization time $t$$$ while full thermal equilibration takes $t$$ and chemical equilibration takes $t$$. At $t$$ the individual momentum-mode occupations are still visibly nonthermal, so the paper is separating bulk thermodynamic quantities from mode-by-mode distributions. The payoff is concrete: if $p=p($$ is already fixed, a hydrodynamic description can be closed much earlier than a strict thermalization argument would allow, which is the puzzle raised by early hydrodynamic behavior in heavy-ion collisions.","feed_headline":"Equation of state locks in before thermal equilibrium","feed_subtitle":"A quark-meson model shows bulk quantities settling about 100 times faster than full thermal equilibration.","key_machinery":"The central machinery is a two-loop two-particle-irreducible (2PI) effective action for a two-flavor quark-meson model, a truncation that resums selected self-energy contributions and retains direct scattering, off-shell, and memory effects. The calculation evolves the quantum fields numerically from specified initial distributions and extracts mode occupation numbers $n$$, mode temperatures $T$$ via a Fermi-Dirac or Bose-Einstein form, the equation-of-state ratio $w=p/$$, a global kinetic temperature $T$$, and species chemical temperatures $T$$ from integrated number densities. The load-bearing mechanism is dephasing: bulk quantities average over many oscillating modes with a dense frequency spectrum, so phase memory is lost quickly, independently of the scattering strength that controls true thermalization. That separation between bulk and mode quantities is what allows the paper to define a prethermalization time and to compare it with $t$$, $t$$, and $t$$.","core_discovery":"The central claim is that prethermalization is a real, quantitatively early stage in a quantum field theory, not merely a classical artifact. For a two-flavor quark-meson model with Yukawa coupling $h$ and quartic coupling $$$, solved in the two-loop 2PI approximation, the paper finds that $w=p/$$ becomes almost time independent after $t$$ of order the inverse thermal mass $m$$; the inset comparing $h=1.0$ with $h=0.5$ shows this time is insensitive to interaction strength. A global kinetic temperature defined by $T$$ with $c$$ fixed by equilibrium, settles on the same short scale and approaches $T$$. By contrast, mode quantities lose initial-condition memory on $t$$ for fermions and scalars, and true thermal equilibration, where all mode temperatures agree, takes $t$$; chemical temperatures for scalars and fermions equalize only on the longer chemical time $t$$. The authors use the in-model relation $T t$$ to estimate $t$$ somewhat below $1$$ for temperatures above about $400$--$500$$, and argue this is consistent with early hydrodynamic behavior.","pith_inferences":["If the bulk-versus-mode separation is generic, other systems with dense frequency spectra and conserved total energy, such as classical lattice gauge theory, ultracold atomic gases, or spin chains, should show an early plateau in their equation of state even when their mode distributions are far from equilibrium.","A direct numerical test would initialize hydrodynamic simulations at $t$$ and at $t$$ and compare final observables; if they agree, that would turn prethermalization into a practical shortcut rather than a correction.","The relation $T t$$ suggests a dimensionless clock that could be checked across couplings, temperatures, and particle content in the same model, and eventually in QCD.","Experimental claims of early thermalization should specify which observable is meant; the paper's logic implies that bulk and mode observables can give different answers without contradiction."],"forward_implications":["A hydrodynamic description of a heavy-ion collision can be closed with $p=p($$ before local thermal equilibrium is established, because the equation of state is already approximately fixed at $t$$.","At temperatures above about $400$--$500$$, the prethermalization estimate is somewhat below $1$$, matching the observed early applicability of hydrodynamics.","Once the scalar and fermion chemical temperatures equalize, their common value coincides with $T$$, so a chemical freeze-out temperature can be compared meaningfully with the equilibrium critical temperature.","Observables that depend on individual momentum-mode occupations can remain out of equilibrium at times when bulk thermodynamic observables already look thermal.","Chemical equilibration remains a late, species-dependent process, so hadron-yield equilibration is not implied by the early constancy of the equation of state."],"supporting_citations":[{"why":"Supplies the two-loop 2PI calculation, numerical method, and mode-temperature definition on which the results rest.","marker":"[6]"},{"why":"Supplies the 2PI far-from-equilibrium framework and the next-to-leading-order $N_f$ expansion used to justify the truncation.","marker":"[2]"},{"why":"Supplies the classical equipartition and kinetic-temperature intuition behind the definition of $T$$.","marker":"[7]"},{"why":"Supplies the dephasing mechanism used to explain why prethermalization is fast and interaction-independent.","marker":"[9]"},{"why":"Supplies the heavy-ion early-hydrodynamics puzzle that motivates the paper's application.","marker":"[1]"},{"why":"Supplies the chemical freeze-out and $T$$ discussion used in the final paragraph.","marker":"[17]"},{"why":"Supplies the QCD equation-of-state behavior above $T_c$ invoked to argue that $w$ can adapt easily.","marker":"[16]"},{"why":"Supplies the proposed QCD extension on which the less-than-$1$$ estimate depends.","marker":"[25]"}],"fun_headline_variants":["Quark-meson model shows prethermalization on short time scales","Equation of state settles before full thermal equilibrium","Kinetic temperature tracks equilibrium early in the quark-meson model","Prethermalization time is insensitive to two Yukawa couplings","Two-loop quark-meson model supports early hydrodynamic behavior"],"cache_read_input_tokens":0,"weakest_assumption_plain":"The load-bearing premise is that the separation of time scales found in the truncated two-loop quark-meson model, together with the numerical relation $T t$$, survives in QCD near $400$--$500$$; the in-model result alone does not establish the less-than-$1$$ heavy-ion estimate.","fun_headline_variants_meta":{"raw":{"variants":["Quark-meson model shows prethermalization on short time scales","Equation of state settles before full thermal equilibrium","Kinetic temperature tracks equilibrium early in the quark-meson model","Prethermalization time is insensitive to two Yukawa couplings","Two-loop quark-meson model supports early hydrodynamic behavior"]},"model":"selfhosted/deepseek-v4.1-flash","effort":"low","attempts":1,"cost_usd":0.0,"raw_usage":{"total_tokens":1595,"prompt_tokens":800,"completion_tokens":795,"prompt_tokens_details":null,"completion_tokens_details":{"reasoning_tokens":728}},"tokens_in":800,"tokens_out":795,"duration_ms":14482,"image_pages":[],"temperature":1.0,"reasoning_budget":50,"reasoning_tokens":728,"transport_provider":"selfhosted","cache_read_input_tokens":0,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-11-13T08:49:53.898392+00:00","model_set":{"reader":"selfhosted/deepseek-v4.1-flash"},"pipeline_version":"pith-pipeline@v1.5.0-archive-selfhost","input_kind":null,"input_chars":null,"input_total_chars":null,"passes_missing":null,"published_standing":null,"published_confidence":null,"objection_audit":null,"run_record":null,"basis":{"kind":"not_recorded","source":null,"chars_given":null,"chars_total":null},"receipt":null,"falsifier":"A controlled first-principles QCD calculation of an expanding quark-gluon plasma that finds $w=p/$$ or $T$$ still drifting by more than about 20% at $1$$, or that yields a substantially different constant in $T t$$, would undercut the paper's heavy-ion extrapolation; a repeat of the same 2PI calculation with a different truncation or coupling finding no early plateau would undercut the in-model claim.","supporting_citations":[{"cited_title":"Chemical Freeze-out and the QCD Phase Transition Temperature","cited_arxiv_id":"nucl-th/0311005","evidence_quote":"Supplies the chemical freeze-out and $T$$ discussion used in the final paragraph."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Supplies the two-loop 2PI calculation, numerical method, and mode-temperature definition on which the results rest."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Supplies the 2PI far-from-equilibrium framework and the next-to-leading-order $N_f$ expansion used to justify the truncation."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Supplies the classical equipartition and kinetic-temperature intuition behind the definition of $T$$."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Supplies the dephasing mechanism used to explain why prethermalization is fast and interaction-independent."},{"cited_title":"Theory Highlights of Quark Matter 2004","cited_arxiv_id":"nucl-th/0403048","evidence_quote":"Supplies the heavy-ion early-hydrodynamics puzzle that motivates the paper's application."},{"cited_title":"Thermodynamics and in-medium hadron properties from lattice QCD","cited_arxiv_id":"hep-lat/0305025","evidence_quote":"Supplies the QCD equation-of-state behavior above $T_c$ invoked to argue that $w$ can adapt easily."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Supplies the proposed QCD extension on which the less-than-$1$$ estimate depends."}],"lane_accuracy":{"status":"not yet measured","lane":{"model":"selfhosted/deepseek-v4.1-flash","input_kind":null},"url":"https://pith.science/accuracy"},"review_version":1}