{"id":"6152e074-5b2d-4560-9ddb-5f836e753262","arxiv_id":"2607.23483","paper_version":1,"verdict":"ACCEPT","confidence":"HIGH","novelty_score":5.5,"correctness_risk":"low","formal_verification":"none","parameter_count":0,"one_line_summary":"Pair antibunching g_pairs^(2)<1 is classically forbidden by Cauchy–Schwarz on the Glauber–Sudarshan P-function and is measurable by fourfold coincidences.","lead":"The paper defines a pair-level correlation g_pairs^(2) from the biphoton operator and proves classical fields cannot go below 1. That makes pair antibunching a clean nonclassicality test that standard heralded g^(2) does not directly give.","discovery_kind":"new_method","skeptic_critique":{"model":"moonshotai/kimi-k3","headline":"No significant objection identified. The classical bound g_pairs^(2)≥1 is a one-line Cauchy–Schwarz/variance statement about the classical pair amplitude |αβ|², and the worked examples (TMSV, truncations) reproduce correctly under independent recomputation.","rationale":"The reader's verdict (ACCEPT, HIGH confidence, low correctness risk) matches my independent read. The reader's weakest_assumption — that positive-P classicality is the right notion for the witness and that the witness does not certify correlated pair generation — is exactly the correct soft spot, and crucially it is disclosed in the paper's own Conclusion rather than hidden. Every quantitative claim I could check by hand reproduced: the variance form of the classical bound, the TMSV moments via geometric factorial moments, the N=2 limit 12/25 at n̄=1, the N=3/N=4 crossover structure of Fig. 1, and loss invariance (which in fact holds mode-by-mode, stronger than stated). The paper is short, parameter-free, and self-contained; there is no derivation long enough to hide an error in, and no empirical claim that could fail outside the math. The only residual concerns are interpretive framing (the non-bosonic nature of P and the product-state caveat), both acknowledged in the text. This is a case where the honest stress-test outcome is a non-finding, and the concrete test above is a worthwhile confirmatory verification rather than a probe of a suspected weakness.","tokens_in":9170,"tokens_out":4905,"duration_ms":165023,"concrete_test":"Two-part numerical verification: (a) Monte Carlo sample several explicit positive P-functions (e.g., Gaussian mixtures over α,β, and classically correlated thermal-type distributions) and compute g_pairs^(2) from Eq. (4); if any instance yields <1, the CS step is being misapplied (expected: all ≥1). (b) Independently recompute Eqs. (A5)–(A8) and (B2) by direct summation over number states truncated at n=50 instead of Wick contractions; agreement to 1e-10 would confirm the Gaussian moment factoring. As a sanity check on the disclosed caveat, verify g_pairs^(2)(|1⟩⊗|1⟩)=0, confirming the witness fires on product nonclassicality.","verdict_should_be":"UNCHANGED","load_bearing_attack":"I checked the load-bearing steps independently and could not find a flaw that threatens the central claim. (1) The classical bound: with P(α,β)≥0, numerator = ∫P|αβ|⁴ and denominator = (∫P|αβ|²)², so g_pairs^(2)−1 = Var(|z|²)/⟨|z|²⟩² ≥ 0. This is exact and assumption-minimal. (2) TMSV (App. A): ⟨n²⟩ = n̄(2n̄+1) for the geometric pair distribution, and expanding n²(n−1)² = F4+4F3+2F2 with geometric factorial moments ⟨F_k⟩=k!n̄^k gives 4n̄²(6n̄²+6n̄+1), matching (A7); limits g→4 (n̄→0) and g→6 (n̄→∞) check out, as does g>4 for all n̄>0. (3) Truncations (App. B): N=2 gives 4(1+x+x²)/(1+4x)²→12/25<1 at x→1 with n̄→1; spot-checking (B2) at x=1 for N=3 yields 160/196≈0.82 (antibunched) and N=4 yields 920/900≈1.02 (bunched), consistent with Fig. 1. (4) Loss invariance: numerator scales as (η_aη_b)² and denominator likewise, so the ratio is invariant even under unequal losses — the paper's claim is conservative. The genuinely soft spots are interpretive, not mathematical, and both are disclosed by the authors: (a) the witness certifies nonclassicality of the joint field, not correlated pair generation — e.g., the product Fock state |1,1⟩ has g_pairs^(2)=0 with no intermode correlation (Conclusion paragraph acknowledges this); (b) P=ab is not a bosonic mode (footnote 7), so for perfectly number-correlated states the quantity equals ⟨[n(n−1)]²⟩/⟨n²⟩² rather than the g^(2) of a literal pair-number operator, which is why TMSV yields 4 rather than the thermal value 2 — the formalism is self-consistent but the \"pair-event statistics\" language is an interpretive overlay. Neither issue undermines the stated theorem.","agreement_with_reader":"agree"},"referee_report":{"model":"moonshotai/kimi-k3","summary":"The authors define a \"pair\" second-order correlation function g_pairs^(2) = ⟨(P†)²P²⟩/⟨P†P⟩² built from the biphoton operator P† = a†b†, and prove via Cauchy–Schwarz that any two-mode field with a positive Glauber–Sudarshan P-function satisfies g_pairs^(2) ≥ 1 (Eqs. 1–6). Pair antibunching (g_pairs^(2) < 1) is therefore a nonclassicality witness accessible through twofold/fourfold coincidence counting, with no phase reference and with invariance under linear loss. Worked examples: product coherent states saturate the bound (g=1); TMSV gives g = 4(6n̄²+6n̄+1)/(2n̄+1)² (App. A), bunched for all nonzero squeezing with limits 4 (weak) and 6 (strong); truncated TMSV states (App. B) share the weak-limit value 4 for N≥2 but N=2,3 cross into antibunching at large λ (N=2 → 12/25). A comparison with heralded g_s^(2) (Fig. 2) shows the two witnesses probe complementary statistics.","tokens_in":9703,"tokens_out":4471,"duration_ms":279222,"significance":"The result is a clean, parameter-free nonclassicality bound: the proof is a one-line application of Cauchy–Schwarz to the classical pair amplitude |z|² = |αβ|², with no fitted constants or normalization tricks, and the loss-invariance argument is exact (numerator and denominator both scale as (η_a η_b)², so the ratio is in fact invariant even under unequal losses — the paper's \"uniform loss\" claim is conservative). I independently verified the load-bearing algebra: the TMSV eight-operator moment 4n̄²(6n̄²+6n̄+1) (Eq. A7) follows from expanding n²(n−1)² in factorial moments of the geometric pair distribution; the N=2 truncation gives 4(1+x+x²)/(1+4x)² → 12/25 < 1 (Eq. B8); and spot checks of Eq. (B2) at x=1 reproduce the N=3 (antibunched) and N=4 (bunched) crossover structure of Fig. 1. The witness is experimentally realistic with existing coincidence-counting technology and fills a genuine gap between single-mode g^(2) and heralded/cross-correlation criteria. The finding that arbitrarily weak TMSV is pair-bunched (g→4) while its heralded signal antibunches is a useful, counterintuitive clarification for the biphoton-source community.","major_comments":[],"minor_comments":[{"comment":"The abstract states invariance 'under uniform loss', while the Results section states the bound holds 'under arbitrary linear attenuation', and Fig. 2 in fact demonstrates invariance under idler-only loss (η = 1 vs 0.5 curves overlap). Since the ratio is invariant under arbitrary unequal losses ((η_a η_b)² cancels), the abstract undersells the result; harmonize the three statements.","section":"Abstract / Results vs. Fig. 2"},{"comment":"Because P = ab is not a canonical bosonic mode, g_pairs^(2) is not literally the g^(2) of a pair-number operator; for perfectly number-correlated states it equals ⟨[n(n−1)]²⟩/⟨n²⟩², which is why TMSV yields 4 rather than the thermal value 2. Footnote 7 acknowledges the commutation issue, but a sentence in the main text giving the number-correlated form would preempt confusion about the value 4.","section":"Discussion / Footnote 7"},{"comment":"Table I labels g_pairs^(2) < 1 an 'unambiguous signature of nonclassical pair statistics', while the Conclusion correctly notes that product states with a nonclassical single-mode factor (e.g., |1,1⟩, for which g_pairs^(2) = 0) also violate the bound, so the witness certifies nonclassicality of the joint field, not correlated pair generation. Align the Table I caption wording with the Conclusion caveat.","section":"Table I and Conclusion"},{"comment":"The novelty claim ('an analogous antibunching criterion for genuine biphoton emission has not been established') would benefit from situating the witness relative to existing higher-order two-mode nonclassicality criteria (e.g., classical Cauchy–Schwarz bounds on signal–idler cross-correlations, and prior work on biphoton/pair correlation functions), so readers can see precisely what is new.","section":"Introduction"},{"comment":"The caption renders as 'heralded g(2)s, =1' etc.; the η symbols appear to have been dropped in compilation. Also state explicitly that the two g_pairs^(2) curves coincide exactly (loss invariance), rather than merely 'overlapping'.","section":"Fig. 2 caption"},{"comment":"Typo: 'H.-S. Goan. is also grateful' — stray period.","section":"Acknowledgments"}],"recommendation":"accept","confidential_remarks":"The central bound is mathematically elementary (a single Cauchy–Schwarz step), so the contribution is conceptual rather than technical; I nonetheless find it correct, clearly written, and useful for the biphoton-source community, and the examples are verified. The authors are commendably transparent about the witness's limitations (joint-field nonclassicality rather than certified pair generation; non-bosonic nature of P). I could not locate prior art establishing this exact pair-level bound, but the editor may wish to confirm the novelty claim against the literature on higher-order two-mode correlation witnesses."},"author_rebuttal":null,"desk_editor":{"model":"grok-4.5","letter":"The load-bearing result is simple and correct: for any two-mode field with a positive P-function, g_pairs^(2) = ⟨|z|^4⟩/⟨|z|^2⟩² ≥ 1 by Cauchy–Schwarz on z=αβ, so pair antibunching is classically forbidden. Appendices A–B check out under independent recomputation—TMSV gives 4(6n̄²+6n̄+1)/(2n̄+1)² with limits 4 and 6, and the N=2 truncation really does go to 12/25. Loss invariance of the ratio is immediate.\n\nWhat is actually new is the packaging: define the second-order correlator through P=ab, prove the classical bound, and show that even arbitrarily weak TMSV remains pair-bunched while certain finite truncations cross into antibunching. The contrast with heralded g_s^(2) is the practical payoff—conditional single-photon antibunching and intrinsic pair statistics answer different questions, and the paper makes that distinction cleanly. Experimental accessibility via fourfold/twofold coincidences is standard and plausible; no phase reference or tomography needed.\n\nSoft spots are interpretive and already flagged by the authors. P is not a bosonic mode, so for number-correlated states this is really ⟨[n(n−1)]²⟩/⟨n²⟩² rather than a literal pair-number g^(2)—which is why TMSV yields 4, not the thermal 2. The witness certifies joint-field nonclassicality, not correlated biphoton generation; product states like |1,1⟩ also give zero. Neither issue breaks the theorem. Novelty is moderate (intensity correlations + CS bounds are old tools), significance is subfield-scoped, and there is no new data—just closed-form theory.\n\nThis is for people who characterize SPDC/FWM sources and want a loss-tolerant complement to heralded g^(2). Math and citation pattern look solid. I would send it to peer review; a serious referee can handle the interpretive caveats in revision. Worth engaging if you work on biphoton diagnostics.","headline":"Clean Cauchy–Schwarz bound on pair-operator correlations; the useful surprise is that weak TMSV still pair-bunches (g→4), while some truncations antibunch.","tokens_in":10734,"tokens_out":547,"would_cite":true,"duration_ms":20321,"reading_group":"maybe","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":[],"pacs":[],"model":"grok-4.5","headline":"Pair antibunching below one is classically forbidden and certifies nonclassical two-mode light.","keywords":["pair correlation","photon antibunching","two-mode squeezed vacuum","nonclassicality witness","Glauber–Sudarshan P-function","biphoton","coincidence measurements","heralded g(2)"],"falsifier":"Measure fourfold and twofold signal–idler coincidences on a verified classical two-mode source (or a product coherent state) and check whether g_pairs^(2) stays ≥ 1; or prepare a truncated pair state |0,0⟩ + √p |1,1⟩ and verify that g_pairs^(2) approaches 0.","tokens_in":10365,"feed_emoji":"🔬","tokens_out":990,"duration_ms":17996,"temperature":0.7,"pith_summary":"This paper defines a second-order correlation for photon pairs, built from the biphoton operator that creates one photon in each of two modes. That single number tells whether pair-generation events bunch, arrive independently, or antibunch. Using a classical inequality on the pair amplitude, the authors prove that every classical two-mode field must score at least one, so any measured value below one is an unambiguous nonclassical signature. They show that familiar two-mode squeezed vacuum states actually bunch at the pair level—even when very weak—while truncated pair states can antibunch, and they contrast this with the more familiar heralded single-photon correlation. The quantity is measured with ordinary coincidence counting, needs no phase reference, and is unchanged by uniform loss, so it offers a practical witness of nonclassical pair emission complementary to existing tests.","feed_headline":"Pair antibunching below 1 is classically forbidden","feed_subtitle":"A coincidence-based witness shows weak squeezed vacuum still bunches at the pair level","key_machinery":"The pair correlation g_pairs^(2) defined through the biphoton operator P† = a† b†. For classical fields it reduces to ⟨|z|⁴⟩ / ⟨|z|²⟩² with z = αβ; Cauchy–Schwarz on |z|² then forces the bound ≥ 1, which is loss-invariant and measured by fourfold versus twofold coincidences.","core_discovery":"The pair second-order correlation g_pairs^(2) = ⟨(P†)² P²⟩ / ⟨P† P⟩², with P† = a† b†, satisfies g_pairs^(2) ≥ 1 for every classical two-mode field with a positive Glauber–Sudarshan P-function. Therefore observed pair antibunching g_pairs^(2) < 1 is classically forbidden and certifies nonclassicality of the joint field. Even arbitrarily weak two-mode squeezed vacuum states yield pair bunching (approaching 4), while finite truncations can cross into antibunching.","pith_inferences":["Source characterization tables for SPDC and four-wave-mixing could routinely list g_pairs^(2) next to heralded g_s^(2) to separate multipair fluctuation from conditional single-photon purity.","Engineering a bright source with g_pairs^(2) < 1 would require deliberate multipair truncation or filtering, not only lower pump power.","Because product nonclassical states can also go below 1, combining g_pairs^(2) with a cross-correlation or entanglement test would separate joint-field nonclassicality from true pair generation."],"forward_implications":["Pair antibunching g_pairs^(2) < 1 becomes a standard, loss-tolerant nonclassicality witness for biphoton platforms, alongside heralded g_s^(2).","Weak two-mode squeezed vacuum sources must be reported as pair-bunched (limit 4), not assumed pair-antibunched from weak excitation alone.","State truncation or multipair suppression, not merely low pump power, is required if the goal is pair antibunching.","Standard coincidence hardware already suffices; no phase lock or tomography is needed to deploy the witness."],"fun_headline_variants":["Pair antibunching g_pairs^(2)<1 is classically forbidden","Cauchy-Schwarz bars classical fields from pair antibunching","Weak two-mode squeezed vacuum still shows pair bunching","g_pairs^(2) witnesses nonclassicality via pair events alone","Pair-level g^(2) certifies nonclassical joint fields"],"cache_read_input_tokens":128,"weakest_assumption_plain":"That a positive Glauber–Sudarshan P-function is the right definition of classical for this test, so beating the bound certifies nonclassicality of the joint field—even though product states with one nonclassical mode can also violate it and the witness alone does not prove correlated biphoton generation.","fun_headline_variants_meta":{"raw":{"variants":["Pair antibunching g_pairs^(2)<1 is classically forbidden","Cauchy-Schwarz bars classical fields from pair antibunching","Weak two-mode squeezed vacuum still shows pair bunching","g_pairs^(2) witnesses nonclassicality via pair events alone","Pair-level g^(2) certifies nonclassical joint fields"]},"model":"grok-4.5","effort":"low","cost_usd":0.003611,"raw_usage":{"total_tokens":1281,"prompt_tokens":910,"num_sources_used":0,"completion_tokens":73,"cost_in_usd_ticks":36108000,"prompt_tokens_details":{"text_tokens":910,"audio_tokens":0,"image_tokens":0,"cached_tokens":256},"completion_tokens_details":{"audio_tokens":0,"reasoning_tokens":298,"accepted_prediction_tokens":0,"rejected_prediction_tokens":0}},"tokens_in":910,"tokens_out":73,"duration_ms":5412,"temperature":1.0,"reasoning_tokens":298,"cache_read_input_tokens":256,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-07-30T21:07:30.897724+00:00","model_set":{"reader":"grok-4.5"},"falsifier":"Measure fourfold and twofold signal–idler coincidences on a verified classical two-mode source (or a product coherent state) and check whether g_pairs^(2) stays ≥ 1; or prepare a truncated pair state |0,0⟩ + √p |1,1⟩ and verify that g_pairs^(2) approaches 0.","supporting_citations":[],"review_version":1}