{"id":"2ba773a3-a2aa-4259-b11a-4a60c4db26b6","arxiv_id":"2411.10155","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":4.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":0,"one_line_summary":"Half-wormholes restore factorization in the one-time-point N=1 supersymmetric SYK model, and both wormholes and half-wormholes break supersymmetry.","lead":"The paper shows that in a simplified version of the N=1 supersymmetric SYK model, adding half-wormhole contributions to the non-averaged squared partition function restores factorization at large N. It also finds that wormholes and half-wormholes both break supersymmetry completely.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Central factorization claim rests on the unproven large-N counting α=4 behind ⟨Z⁴⟩=3⟨Z²⟩² (Eq. 92); the Appendix C argument is a heuristic assumption, not a calculation.","rationale":"The paper's strongest claim — that including half-wormholes restores factorization, Z² ≈ ⟨Z²⟩ + Φ(0) — is made rigorous only if the error cancellation (55) holds. That cancellation depends on the precise value of α in Eq. (91), which determines ⟨Z⁴⟩=3⟨Z²⟩². The derivation of α=4 is the weakest point: the triple sum in Eq. (90) is not evaluated, the plot in Fig. 1 covers only one symmetric mixed configuration, and the assertion that all other terms are subleading is an assumption (footnote 4). This is exactly the reader's weakest_assumption, and it is load-bearing because any change in α changes the coefficient of ⟨Z²⟩² in (55) and destroys the factorization argument. I agree with the reader's conditional verdict: the logic is plausible and follows a well-established framework, but the central claim is not fully secured until the counting is checked. The SUSY-breaking observation in Section 4 is a separate, more straightforward argument. The exact ⟨Z²⟩ result is checked by two independent methods, which gives credit to the overall approach; the concern is isolated to the fourth-moment counting.","tokens_in":15649,"tokens_out":10754,"duration_ms":96882,"concrete_test":"Evaluate the triple sum in Eq. (90) numerically for N=8,16,32,64 and verify that its ratio to the one-pair contribution K1|k_LR=k_L'R'=N/2 decays exponentially in N; then repeat the check for the remaining mixed derivative configurations in Eq. (83), including cases with two or three of l1,l2,l3 nonzero, and for the other 63 terms in the sum (77). If any contribution is not exponentially suppressed, α≠4 and Eq. (92) fails.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The restoration of factorization, Eq. (51), is established through the mean-squared-error cancellation (55), which requires ⟨Z⁴⟩=3⟨Z²⟩², ⟨Φ(0)²⟩=2⟨Z²⟩², and −2⟨Z²Φ(0)⟩ contributing −4⟨Z²⟩². The first of these is derived in Appendix C only up to an undetermined coefficient α in Eq. (91). The paper sets α=4 by asserting that leading large-N contributions come only from terms with derivatives with respect to two distinct G's, and that all other terms — including mixed derivative terms from Eq. (83) and the remaining 63 terms in the sum (77) — are exponentially suppressed. The only evidence for this suppression is Fig. 1, which plots a single symmetric mixed configuration (k_LR=k_L'R'=k_LR'=k_RL'=N/4) and is not a closed-form evaluation of the triple sum (90). Footnote 4 records that mixed products were observed not to contribute in a previous computation, but that computation is not shown and does not cover all cases. If any unsuppressed mixed configuration exists, α≠4, Eq. (92) fails, and the combination in (55) does not vanish, so the claimed restoration of factorization is not established.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper studies the N=1 supersymmetric SYK model with time reduced to a point. The ensemble average of the partition function squared is computed exactly and reproduced by a saddle-point analysis after a contour deformation; the saddle points are interpreted as wormholes. The fourth moment is then analyzed to identify half-wormhole contributions to the non-averaged Z^2, with the claim that including half-wormholes restores factorization. It is also argued that both wormhole and half-wormhole saddle configurations break supersymmetry completely.","tokens_in":15919,"tokens_out":17606,"duration_ms":137824,"significance":"If the central claim holds, the paper extends the half-wormhole/factorization program to a supersymmetric version of SYK, which is a natural and nontrivial generalization of [21]. The exact ⟨Z^2⟩ computation with the one-loop determinant matching is a concrete technical achievement. The supersymmetry-breaking observation is interesting and potentially relevant for holographic applications. However, the factorization-restoring argument depends on an unproved large-N counting assumption, so the central claim is conditional.","major_comments":[{"comment":"The derivation of ⟨Z^4⟩ = 3⟨Z^2⟩^2 (Eq. (92)) rests on the unproven assumption that α=4. The triple sum in Eq. (90) is not evaluated; Fig. 1 plots only the symmetric mixed configuration with kLR=kL'R'=kLR'=kRL'=N/4 and does not establish suppression of all other mixed configurations. Footnote 4 asserts that mixed products did not contribute in a previous computation, but that computation is not shown and does not cover all terms in (77). Since Eq. (55) cancels the error only when ⟨Z^4⟩ = 3⟨Z^2⟩^2, the factorization restoration claim is conditional on this counting. Please provide a closed-form evaluation of the sum (90) or a rigorous large-N argument that all contributions beyond the three pairing saddles (42) are subleading.","section":"Appendix C, Eq. (91)"},{"comment":"The values ⟨Φ(0)^2⟩ = 2⟨Z^2⟩^2 and ⟨Z^2Φ(0)⟩ = 2⟨Z^2⟩^2 are asserted through the pairing-counting statement 'for the last two terms only two remain'. No explicit computation of these averaged products is presented. If the counting is affected by the same unsettled issue as the α=4 assumption, the cancellation in (55) fails. Please provide the explicit expressions or a controlled derivation for these two correlation functions.","section":"Section 3.2, Eq. (55)"},{"comment":"The constant term in L1 appears to be incorrect. Evaluating (22) at the saddle (26) with m=0 gives L1 = -N + (3N/4) log 3 + (N/2) log J - N log 2, whereas (29) gives -N + (N/2) log(3^{3/2}/2) + N log sqrt(J) = -N + (3N/4) log 3 + (N/2) log J - (N/2) log 2. The displayed L1 therefore yields e^{L1} a factor 2^{N/2} too large, and the sum over the four saddle points would not reproduce Eq. (16). Please correct Eq. (29) and verify that the saddle-point sum, including one-loop determinants, reproduces Eq. (16) with the corrected constant.","section":"Section 2.2, Eq. (29)"}],"minor_comments":[{"comment":"The caption does not define the base of the logarithm or the exact ratio plotted (including N-dependent prefactors); please make the plot reproducible.","section":"Figure 1"},{"comment":"The phrase 'the previous computation' is vague; please give a reference or an appendix where that computation is shown.","section":"Footnote 4"},{"comment":"The notation 'the ψψ label is suppressed on the rhs' is confusing; please define σαβ and gαβ explicitly as (σψψ)αβ and (gψψ)αβ.","section":"Section 2.3, after Eq. (41)"},{"comment":"The concluding remark that a self-averaging region could not be depicted is an honest limitation, but the main text would benefit from a one-sentence recap of how this limitation affects the comparison with [21].","section":"Section 5"}],"recommendation":"major_revision","confidential_remarks":"The manuscript is within the scope of the journal and the topic is timely. The main technical gap is the unproven counting behind Eqs. (91)-(92) and the corresponding correlation functions in Eq. (55). If the authors can supply the missing evaluation or a rigorous argument, the paper would be a solid contribution. I see no grounds for rejection, but the central claim is currently conditional on an unverified assumption."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Short version: this is the first half-wormhole analysis for the N=1 supersymmetric SYK model at one time point, and the SUSY-breaking observation is clean and correct. The main factorization claim, though, is only as solid as an unproven large-N counting assumption in Appendix C, so treat it as conditional rather than established.\n\nWhat the paper does well: the exact computation of ⟨Z²⟩ in Section 2 is careful, and the saddle-point match including one-loop determinants is genuine evidence that the wormhole interpretation is right. The SUSY-breaking argument in Section 4 is simple and convincing: a nonzero G_ψψ saddle forces all ϵ_α to zero, so wormholes and half-wormholes both completely break the N=1 supersymmetry. That observation will be useful to people working on supersymmetric SYK and JT supergravity. The citation pattern is fine; the paper builds directly on Saad-Shenker-Stanford-Yao and cites the surrounding literature.\n\nThe soft spot is the α=4 counting in Appendix C. Equation (91) leaves α undetermined, and the argument that only two-derivative terms contribute at leading order is heuristic. The triple sum in (90) is not evaluated; Fig. 1 only shows one symmetric mixed configuration, and footnote 4 refers to a computation that isn't shown. If an unsuppressed mixed configuration exists, then ⟨Z⁴⟩≠3⟨Z²⟩² and the cancellation in (55) fails. This is a real gap, not a nitpick. The authors are honest about it — they say they 'argue' and 'assume' — but honesty doesn't close it. The relation ⟨Z⁴⟩=3⟨Z²⟩² is consistent with a Gaussian Z, and the plotted suppression is suggestive, so I wouldn't bet against the result, but it is not yet demonstrated.\n\nA minor limitation: factorization restoration is only claimed at saddle-point values of σ_ψψ, not for the full self-averaging region; the authors acknowledge this. That's acceptable for what they set out to do.\n\nWho this is for: people working on wormholes without averaging, SYK variants, and supersymmetric JT gravity. It deserves a serious referee. My recommendation: send it to peer review, and ask the authors to close the α=4 gap — either by evaluating the triple sum in a large-N saddle approximation or by providing a direct fixed-coupling check along the lines of Mukhametzhanov. If they close it, this is a solid paper. As it stands, the main result is conditional.","headline":"First half-wormhole computation for N=1 SUSY SYK, with a solid ⟨Z²⟩ calculation and a neat SUSY-breaking observation, but the factorization claim depends on an unproven counting step in Appendix C.","tokens_in":16438,"tokens_out":3104,"would_cite":true,"duration_ms":26788,"reading_group":"yes","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":[],"pacs":[],"model":"deepseek-v4-flash","headline":"Half-wormholes restore factorization in the supersymmetric SYK model at one time point.","keywords":["half-wormholes","supersymmetric SYK model","factorisation problem","wormhole saddles","large N limit","supersymmetry breaking","random couplings","collective fields"],"falsifier":"Evaluate the unresolved triple sum below Eq. (90) exactly or numerically to large $N$; if the configurations with $k_{LR}=k_{L'R'}=k_{LR'}=k_{RL'}=N/4$ are not exponentially suppressed relative to the two-pair saddle, then $\\langle Z^4\\rangle=3\\langle Z^2\\rangle^2$ fails and the error cancellation in (55) collapses.","tokens_in":15421,"feed_emoji":"🪱","tokens_out":10612,"duration_ms":93756,"temperature":0.7,"pith_summary":"This paper tries to establish that the factorization puzzle in the $N=1$ supersymmetric SYK model is cured by half-wormholes, exactly as it is in the ordinary SYK model. In the simplified setting where time is collapsed to a single instant, the authors show that the non-averaged square of the partition function is reproduced at large $N$ by adding a half-wormhole contribution to the averaged wormhole result: $Z^2\\approx \\langle Z^2\\rangle+\\Phi(0)$, with the variance of the error canceling. The same saddle-point analysis yields $\\langle Z^4\\rangle=3\\langle Z^2\\rangle^2$, so the partition function is effectively Gaussian. A second, independent finding is that wormhole and half-wormhole saddles each force all supersymmetry parameters to zero, breaking the $N=1$ supersymmetry completely.","feed_headline":"Half-wormholes restore factorization in supersymmetric SYK","feed_subtitle":"Half-wormhole saddles make the non-averaged partition function factorize; both saddle types break supersymmetry.","key_machinery":"The engine of the argument is the half-wormhole saddle: a collective-field configuration that contributes to the fourth moment $\\langle Z^4\\rangle$ through one of the three pairings of four boundaries into two wormholes, but contributes nothing to the averaged second moment $\\langle Z^2\\rangle$, so that adding it to the non-averaged expression restores factorization. Technically, the paper rewrites $Z^2$ as an integral over a single auxiliary field $\\Sigma$ and studies the integrand $\\Phi(\\Sigma)$; at $\\Sigma=0$, $\\langle\\Phi(0)\\rangle=0$ while $\\langle\\Phi(0)^2\\rangle=2\\langle Z^2\\rangle^2$, which identifies the half-wormhole region. Wormhole saddles are the non-trivial solutions of the saddle-point equations after contour deformation, solved in Section 2.2.","core_discovery":"For the $\\mathcal{N}=1$ supersymmetric SYK model reduced to a single time point, the paper's central claim is that factorization at fixed couplings is restored in the large $N$ limit by including half-wormhole saddle points. Concretely, $Z^2\\approx \\langle Z^2\\rangle+\\Phi(0)$, where $\\Phi(0)$ is the half-wormhole contribution: it has zero average but nonzero second moment, and the mean squared error vanishes because the wormhole-pairing contributions to $\\langle Z^4\\rangle$, $\\langle Z^2\\rangle^2$, $\\langle \\Phi(0)^2\\rangle$, and $\\langle Z^2\\Phi(0)\\rangle$ count as $3, 1, 2,$ and $4$, respectively, giving $(3-1+2-4)\\langle Z^2\\rangle^2=0$. Along the way the paper establishes $\\langle Z^4\\rangle=3\\langle Z^2\\rangle^2$, consistent with a Gaussian distribution for $Z$. It also shows that the nonzero fermion-bilinear saddle value $G^{LR}_{\\psi\\psi}$ forces, through the composite-field transformation rules, every supersymmetry parameter $\\epsilon_\\alpha$ to vanish; hence wormholes and half-wormholes both break supersymmetry completely.","pith_inferences":["Beyond the paper, the same half-wormhole mechanism is testable at fixed couplings: a direct evaluation of $Z^2$ in the one-time-point $N=1$ model, analogous to the ordinary SYK computation, should reproduce $Z^2\\approx \\langle Z^2\\rangle+\\Phi(0)$.","Beyond the paper, the fact that both wormhole and half-wormhole saddles kill every supersymmetry parameter suggests that a bulk dual such as supersymmetric JT gravity should contain SUSY-breaking saddles with the same pairing structure; checking the $N=2$ or complex SYK variants would show how generic this is.","Beyond the paper, the one-time-point setting leaves open whether the restoration persists once time dependence and the Schwarzian mode are included; if the half-wormhole contribution is purely a saddle-counting effect of the instant reduction, the factorization story would look different in the full model."],"forward_implications":["In the one-time-point $N=1$ supersymmetric SYK model, the non-averaged $Z^2$ factorizes at large $N$ once half-wormholes are included: $Z^2\\approx \\langle Z^2\\rangle+\\Phi(0)$ and the mean squared error vanishes by the pairing count $(3-1+2-4)\\langle Z^2\\rangle^2=0$.","The fourth moment satisfies $\\langle Z^4\\rangle=3\\langle Z^2\\rangle^2$, so at leading order $Z$ is Gaussian-distributed even though the couplings are fixed.","Both wormhole and half-wormhole configurations force all supersymmetry parameters to zero, so these saddles break $N=1$ supersymmetry completely.","The averaged moments are nonzero only when $N$ is a multiple of four, tying the wormhole story to that divisibility condition."],"supporting_citations":[{"why":"Introduces half-wormhole saddle points as contributions to the fourth moment but not the second moment, the construction this paper transplants to the $N=1$ model.","marker":"[21]"},{"why":"Provides the direct fixed-coupling computation in ordinary SYK that previously confirmed half-wormholes restore factorization, the comparison point for the present argument.","marker":"[22]"},{"why":"Supplies the argument the paper repeats to conclude that including half-wormholes suffices to reproduce the non-averaged $Z^2$.","marker":"[32]"},{"why":"Defines the $N=1$ supersymmetric SYK model whose supercharge and Lagrangian the paper uses.","marker":"[35]"}],"fun_headline_variants":["Half-wormholes restore factorization and break SUSY in SYK","Half-wormholes fix factorization, break supersymmetry in SYK","In SYK, half-wormholes restore factorization, break SUSY","Half-wormhole saddles: factorize SYK, break supersymmetry"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The argument leans on the assumption, stated in footnote 4, that in the large-$N$ limit only the three wormhole-pairing saddles contribute to the fourth moment and that all other contributions, including mixed derivative terms, are exponentially suppressed; the decisive triple sum is plotted, not evaluated.","fun_headline_variants_meta":{"raw":{"variants":["Half-wormholes restore factorization and break SUSY in SYK","Half-wormholes fix factorization, break supersymmetry in SYK","In SYK, half-wormholes restore factorization, break SUSY","Half-wormhole saddles: factorize SYK, break supersymmetry"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000639,"raw_usage":{"total_tokens":2897,"prompt_tokens":856,"completion_tokens":2041,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":472,"completion_tokens_details":{"reasoning_tokens":1961}},"tokens_in":472,"tokens_out":2041,"duration_ms":14407,"temperature":1.0,"reasoning_tokens":1961,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-12T19:54:03.132351+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Evaluate the unresolved triple sum below Eq. (90) exactly or numerically to large $N$; if the configurations with $k_{LR}=k_{L'R'}=k_{LR'}=k_{RL'}=N/4$ are not exponentially suppressed relative to the two-pair saddle, then $\\langle Z^4\\rangle=3\\langle Z^2\\rangle^2$ fails and the error cancellation in (55) collapses.","supporting_citations":[{"cited_title":"Half-Wormholes and Ensemble Averages","cited_arxiv_id":"2205.01288","evidence_quote":"Supplies the argument the paper repeats to conclude that including half-wormholes suffices to reproduce the non-averaged $Z^2$."}],"review_version":1}