{"id":"1bd76cfd-a9f9-4a31-8219-f19a83b133ed","arxiv_id":"2608.21823","paper_version":1,"verdict":"ACCEPT","confidence":"MODERATE","novelty_score":8.0,"correctness_risk":"low","formal_verification":"none","parameter_count":0,"one_line_summary":"The endpoint orthonormal Strichartz estimate holds in the abstract Keel-Tao framework, resolving the Frank-Lewin-Lieb-Seiringer conjecture for the free Schrodinger equation in dimensions d at least 2.","lead":"This paper proves a new endpoint inequality for orthonormal families of dispersive wave solutions under the same general hypotheses as the classical Keel-Tao theorem. It settles two long-standing conjectures in harmonic analysis and sharpens a one-function Strichartz estimate to the optimal regularity exponent.","discovery_kind":"first_principles","skeptic_critique":{"model":"deepseek-v4-flash","headline":"No significant objection identified; the endpoint orthonormal Strichartz proof is internally consistent under the stated Keel-Tao hypotheses.","rationale":"The reader's weakest_assumption correctly identifies Assumption 1.3 with sigma > 1/2 as the threshold condition: it enters through r = 2 sigma + 1 > 2 in Lemma 4.2 and through the integrability of the kernel in Lemma 3.5, and it is exactly the condition that makes the near-far decomposition summable in Lemma 5.1. I have checked the main proof steps and found no internal inconsistency. The near-part absorption uses a fixed epsilon chosen only from sigma, and the far-part constant's epsilon^{-(r-2)} divergence is harmless because epsilon is not sent to zero. The spectral band argument in Lemma 4.7 is sound: the choice of lambda* is justified by finiteness of the sup, and the Schatten estimate for the m >= 1 bands uses the correct Holder exponents for the trace. The density extension to general V is standard, and the formal integral definition in Remark 1.5 is supported by the Keel-Tao single-function estimate with admissible pair (2p,2q), which follows from the same two assumptions. I therefore agree with the reader's ACCEPT verdict at moderate confidence: the proof is detailed and consistent, and the main risk is the usual one for a long analysis paper without formal verification or independent human checking.","tokens_in":22772,"tokens_out":54231,"duration_ms":459378,"concrete_test":"Worth running: independently re-derive the trace bound in Lemma 4.7(iii) for m >= 1, using |Tr(P0 B P0)| <= N0^{1-2/r} ||B||_{S_{r/2}} with the Schatten-Holder identity 2/r + (r-2)/r = 1, and then verify the summation bound in Lemma 5.1 with the exact definition of w(n,m) for m >= 1, checking that min(w,2^{-m}) <= C epsilon^{1/4} 2^{-m/4} 2^{-n/16} follows from min(a,b) <= sqrt(ab) and w <= epsilon^{1/2} 2^{m/2 - n/8}. If either step fails or produces a logarithmic divergence, the epsilon-absorption in Proposition 5.3 would collapse.","verdict_should_be":"UNCHANGED","load_bearing_attack":"After checking the proof chain, I do not find a load-bearing flaw in the central claim. The key condition is indeed Assumption 1.3 with sigma > 1/2: it is used in Lemma 3.5 to make the separated-time kernel integrable, and it gives r = 2 sigma + 1 > 2 so that the discrete kernel in Lemma 4.2 and the summability in Lemma 5.1 have finite mass. For the Schrodinger application this is satisfied with sigma = d/2 >= 1. The near-far decomposition, the choice of ell(k,k') in (4.1), the spectral band selection in Lemma 4.7, and the epsilon-absorption in Proposition 5.3 are internally consistent; in particular Lemma 5.1's epsilon^{1/4} bound supplies the smallness needed to absorb the near part uniformly in V. The step from dyadic step functions to general V in L^r(I;L^{r/2}(X)) is standard: the operator in Remark 1.5 is well-defined via the Keel-Tao single-function estimate with the admissible pair (2p,2q), and the density argument gives the S^{r,infty} bound. I therefore see no reason to alter the ACCEPT verdict.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper proves endpoint orthonormal Strichartz estimates in the abstract Keel--Tao framework. Under Assumptions 1.2 (energy estimate) and 1.3 (dispersive decay with exponent sigma > 1/2), Theorem 1.4 establishes the restricted-type estimate (1.9) with exponents p=(2sigma+1)/(2sigma), q=(2sigma+1)/(2sigma-1), r=2sigma+1, equivalently the weak-type Schatten estimate (1.10). The proof splits the two-time kernel into near and far parts relative to dyadic level sets of V, selects a good spectral scale via Lemma 4.7, and sums the resulting block estimates in Lemma 5.1 to absorb the near contribution. Applications include the Frank--Lewin--Lieb--Seiringer endpoint conjecture for the free Schrodinger equation in dimensions d >= 2 (Corollary 1.8(ii)), the Bennett--Bez--Gutierrez--Lee transport conjecture (Corollary 1.9), and a sharp Besov refinement for single functions (Corollary 1.11).","tokens_in":22974,"tokens_out":30665,"duration_ms":265171,"significance":"If correct, this is a significant advance: it closes a long-standing endpoint conjecture and shows that the original Keel--Tao assumptions, without additional structural hypotheses on the propagator, are sufficient for the orthonormal endpoint estimate. The paper is carefully structured and largely self-contained: the main theorem is stated with explicit dependence on the constants C0 and B, the key length scale ell(k,k') in (4.1) is explicit, and the proof produces concrete, falsifiable consequences for Schrodinger, transport, and Besov refinements. The near-far and eigenvalue-counting argument is technically dense and merits independent verification, but I checked the pivotal estimates (Lemmas 4.2, 4.6, 4.7, and 5.1) and found the chain internally consistent.","major_comments":[],"minor_comments":[{"comment":"The existence of a spectral scale lambda* satisfying Lemma 4.7(i) is asserted with the phrase 'It is easy to see'; since the function zeta -> zeta^r N_{0,zeta} is only a step function, a one-sentence justification (for example, by taking a sequence approaching the supremum and passing to a point where the value is at least half of it) would make the argument fully explicit.","section":"Lemma 4.7"},{"comment":"The point F in Figure 1 is used in the proof of Corollary 1.6 but is never defined in the text, and the notations (F,A) and (A,F) are used interchangeably. Because the proof relies on F being the intersection of the segment [B,A] with the line p=q, the authors should define F explicitly and use a consistent ordering for the segments.","section":"Section 5.4 and Figure 1"},{"comment":"The bound for m >= 1, namely min(w(n,m), 2^{-m}) <= C epsilon^{1/4} 2^{-m/4} 2^{-n/16}, is stated without the case split that justifies it; adding two or three lines showing the two regimes (epsilon 2^{3n/4-m} <= 1 and > 1) would improve readability.","section":"Lemma 5.1"},{"comment":"The passage from dyadic step functions to general V in L^r(I; L^{r/2}(X)) is described only as a standard density argument; spelling out the completeness of the S^{r,infty} quasi-norm and the identification of the limit operator would make the final step of Theorem 1.4 fully explicit.","section":"Section 5.3"}],"recommendation":"accept","confidential_remarks":"This is a technically dense paper whose central argument is a long near-far decomposition combined with a spectral-scale selection. I verified the key intermediate estimates and found no gap. The claims of novelty relative to [12], [26], and [34] appear accurate. I recommend acceptance; the minor comments above are presentation improvements only."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"The headline result is real: Theorem 1.4 gives the restricted-type endpoint at point A in Figure 1 under exactly the Keel–Tao energy and dispersive hypotheses with σ>1/2. That settles the conjecture Frank–Lewin–Lieb–Seiringer raised, which others had approached only in non-endpoint form. The Besov refinement with optimal κ=2p in Corollary 1.11 is a genuine bonus, not an afterthought, and the transport consequence follows cleanly from the known transfer principle.\n\nWhat the paper does well: the proof is self-contained and the strategy is transparent. The near-far decomposition with the scale ℓ(k,k′) chosen in (4.1) is the technical core, and the division of labor between energy (near) and dispersive (far) estimates is clear. The spectral band selection via the almost-maximal scale λ∗ in Lemma 4.7 is clever, and the ε^{1/4} absorption in Lemma 5.1 supplies the right smallness. I checked the main chain: Assumption 1.3 with σ>1/2 enters exactly where it should, the dyadic step function reduction is standard, and the density argument for general V is routine. The constants are tracked explicitly, which is a plus.\n\nSoft spots are few. The proof of Lemmas 4.6–4.7 and the summation in Lemma 5.1 is dense, and this is where an independent check would be most valuable. I did not find a gap, but the sheer number of indices (n, m, n′, k, k′) and the two competing smallness conditions in (4.11) make it easy for a reader to get lost. The interpolation argument in Corollary 1.6 is handled a little quickly; it is plausible, but the Lorentz-space inclusions could be spelled out more. Remark 5.2 acknowledges that the ε^{1/4} could be improved to ε^{1/2}, which is honest and not an issue.\n\nThe citations look fair. The authors compare with Nguyen, Hoshiya, and Feng–Mondal–Song–Wu explicitly and explain what their theorem adds. The self-citations play no role in the proof. No fitted parameters, no circularity.\n\nThis paper deserves a serious referee. The endpoint resolution has been open for a decade and the abstract formulation is the right level of generality. I would recommend acceptance after a careful referee report focuses on the technical lemmas in Section 4.2.","headline":"This paper proves the long-open endpoint orthonormal Strichartz estimate in the abstract Keel–Tao framework and resolves the FLLS conjecture for the Schrödinger equation in d≥2.","tokens_in":23546,"tokens_out":1233,"would_cite":true,"duration_ms":13932,"reading_group":"yes","serious_thinker":"yes","would_accept_peer_review":true},"pith_extraction":{"msc":["35B45","47B10","42B37","35Q41","35Q49"],"pacs":[],"model":"deepseek-v4-flash","headline":"The paper proves the endpoint orthonormal Strichartz estimate in the full Keel–Tao abstract framework, resolving the Frank–Lewin–Lieb–Seiringer conjecture for the free Schrödinger equation in dimensions d ≥ 2 and the dual conjecture for…","keywords":["orthonormal Strichartz estimates","Keel–Tao framework","restricted-type estimate","Lorentz spaces","Schatten–Lorentz classes","Schrödinger equation","free transport equation","Besov refinement"],"falsifier":"Compute the far-part bound (4.4) at $\\sigma=1/2$, where $r=2$: the discrete kernel $2^{-(r-2)|k|/4}$ becomes identically $1$, so the sum over dyadic levels diverges and the constant in Proposition 5.3 blows up; consistently, the restricted-type estimate (1.6) is known to fail for the free Schrödinger equation in $d=1$, the case $\\sigma=1/2$.","tokens_in":22555,"feed_emoji":"🌀","tokens_out":10487,"duration_ms":88799,"temperature":0.7,"pith_summary":"This paper establishes the endpoint (Lorentz) orthonormal Strichartz estimate for any propagator satisfying the same two abstract hypotheses that appear in Keel and Tao's classical theorem: a uniform $L^2$ bound and a dispersive decay $\\|U(t)U(\\tau)^*\\|_{L^1\\to L^\\infty} \\le C_0 |t-\\tau|^{-\\sigma}$ with $\\sigma>1/2$. This closes the one missing range in the orthonormal theory, resolving the conjecture first raised by Frank, Lewin, Lieb, and Seiringer for the free Schrödinger equation in $d\\ge 2$, answering the dual Bennett–Bez–Gutiérrez–Lee conjecture for the free transport equation, and producing an optimal Besov refinement of the single-function Strichartz estimate. The paper's interest is that orthonormal Strichartz estimates are the many-body version of classical dispersive bounds, and the endpoint had resisted earlier approaches because of a logarithmic divergence at $\\sigma=1/2$.","feed_headline":"Orthonormal Strichartz endpoint estimate proved for d ≥ 2","feed_subtitle":"Restricted-type Schatten bound settles the long-open endpoint conjecture and its transport analogue.","key_machinery":"The proof is carried by a near–far decomposition of the operator $fUU^*f$ (with $f=\\sqrt V$) that splits the double time integral according to whether two time points are separated or close. The novel ingredient is the level-dependent time partition $\\ell(k,k') = \\varepsilon \\lambda 2^{-(k+k')/2} 2^{|k-k'|/4}/B^2$ (equation 4.1), chosen so that the dispersive decay controls the far part through a Hilbert–Schmidt bound (Lemma 4.2) while the energy bound controls the near part (Lemmas 4.4–4.6). The near contribution is reorganized by relative dyadic levels and spectral bands $P_{m,\\lambda}$ of $fUU^*f$; a spectral scale $\\lambda_*$ is selected so that higher bands have geometrically decaying ranks (Lemma 4.7), and Lemma 5.1 sums the block estimates to a small constant that is absorbed into the lower bound $\\lambda_* N_{0,\\lambda_*}$.","core_discovery":"Theorem 1.4 states that, under Assumptions 1.2 and 1.3 with $\\sigma>1/2$, the restricted-type orthonormal Strichartz estimate (1.9) holds at the exponent triple $p=(2\\sigma+1)/(2\\sigma)$, $q=(2\\sigma+1)/(2\\sigma-1)$, $r=2\\sigma+1$, with operator norm $C=C_\\sigma C_0^{2/r} B^{2(r-2)/r}$. By the duality principle this is equivalent to the weak-type Schatten estimate (1.10), namely that the time-integrated operator $\\int_I U(t)^*V(t)U(t)\\,dt$ lies in the Schatten–Lorentz class $S^{r,\\infty}(H)$ with bound in $L^r(I;L^{r/2}(X))$. The proof works on an arbitrary $\\sigma$-finite measure space $X$ and interval $I$, so the result is a genuinely abstract endpoint theorem rather than a property of the Schrödinger equation.","pith_inferences":["At the threshold $\\sigma=1/2$ the proof's summation kernel $2^{-(r-2)|k|/4}$ loses summability, and the Schrödinger case $d=1$ fails; this suggests the near–far decomposition is sharp and no abstract endpoint theorem of this type can hold at $\\sigma=1/2$.","The level-dependent time partition (4.1) is a transferable device: it could be adapted to endpoint orthonormal estimates for wave, Klein–Gordon, or restriction problems wherever a dispersive decay exponent above $1/2$ is available.","The optimal Besov refinement with $\\kappa=2p$ hints at a general principle: orthonormal Strichartz machinery recovers exactly the critical regularity scale for a single function; an analogous sharp refinement for wave equations would be a natural test.","Because only the two Keel–Tao assumptions are used, the theorem should extend to non-Euclidean settings (compact manifolds, Lie groups, graphs) with the same decay; verifying the endpoint there would broaden the resolution well beyond Euclidean space."],"forward_implications":["The conjecture of Frank, Lewin, Lieb, and Seiringer for the free Schrödinger equation in $d\\ge 2$ is settled: the restricted-type estimate (1.6) holds with Lorentz norm $\\ell^{(d+1)/d,1}$.","The Bennett–Bez–Gutiérrez–Lee conjecture for the free transport equation, in its dual form, follows for $d\\ge 2$: the density estimate (1.7) holds.","The single-function Strichartz estimate is refined to the sharp Besov exponent: $\\|e^{it\\Delta}\\varphi\\|_{L^{2p}_t L^{2q}_x} \\lesssim \\|\\varphi\\|_{\\dot B^0_{2,2p}}$, and no exponent $\\kappa<2p$ can replace $2p$.","Any propagator that satisfies the Keel–Tao energy and dispersive hypotheses with $\\sigma>1/2$ inherits the endpoint orthonormal estimate, so the result covers a broad class of dispersive evolutions beyond the Schrödinger equation."],"supporting_citations":[{"why":"Supplies the abstract energy and dispersive hypotheses (Assumptions 1.2 and 1.3) and the classical endpoint theorem that the paper extends.","marker":"[27]"},{"why":"Introduced the orthonormal Strichartz estimate, proved the non-endpoint range, and identified the endpoint with its logarithmic divergence; this is the conjecture the paper resolves.","marker":"[14]"},{"why":"Extended the range up to the endpoint and raised the weak-type Schatten estimate as an open question; also provides the passage to Besov refinements.","marker":"[16]"},{"why":"Explicitly stated the endpoint restricted-type estimate as a conjecture, proved the failure in d=1, and gave the semiclassical link to the transport equation via Proposition 5.1.","marker":"[3]"},{"why":"Conjectured the kinetic transport analogue in dual form, which Corollary 1.9 answers.","marker":"[2]"},{"why":"Earlier abstract orthonormal Strichartz result in a semiclassical setting, used as the comparison baseline showing the missing endpoint.","marker":"[34]"}],"fun_headline_variants":["Endpoint orthonormal Strichartz conjecture resolved","Schatten bound settles endpoint Strichartz question","Keel-Tao orthonormal endpoint finally proven","Orthonormal Strichartz endpoint solved for d≥2"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The argument rests entirely on the decay bound $\\|U(t)U(\\tau)^*\\|_{L^1\\to L^\\infty} \\lesssim |t-\\tau|^{-\\sigma}$ with $\\sigma>1/2$; at $\\sigma=1/2$ the summation over dyadic levels ceases to converge and the estimate is known to fail for the Schrödinger equation in $d=1$.","fun_headline_variants_meta":{"raw":{"variants":["Endpoint orthonormal Strichartz conjecture resolved","Schatten bound settles endpoint Strichartz question","Keel-Tao orthonormal endpoint finally proven","Orthonormal Strichartz endpoint solved for d≥2"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.001005,"raw_usage":{"total_tokens":4207,"prompt_tokens":857,"completion_tokens":3350,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":473,"completion_tokens_details":{"reasoning_tokens":3287}},"tokens_in":473,"tokens_out":3350,"duration_ms":24920,"temperature":1.0,"reasoning_tokens":3287,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-27T18:47:47.877127+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Compute the far-part bound (4.4) at $\\sigma=1/2$, where $r=2$: the discrete kernel $2^{-(r-2)|k|/4}$ becomes identically $1$, so the sum over dyadic levels diverges and the constant in Proposition 5.3 blows up; consistently, the restricted-type estimate (1.6) is known to fail for the free Schrödinger equation in $d=1$, the case $\\sigma=1/2$.","supporting_citations":[{"cited_title":"Keel and T","cited_arxiv_id":null,"evidence_quote":"Supplies the abstract energy and dispersive hypotheses (Assumptions 1.2 and 1.3) and the classical endpoint theorem that the paper extends."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Introduced the orthonormal Strichartz estimate, proved the non-endpoint range, and identified the endpoint with its logarithmic divergence; this is the conjecture the paper resolves."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Extended the range up to the endpoint and raised the weak-type Schatten estimate as an open question; also provides the passage to Besov refinements."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Explicitly stated the endpoint restricted-type estimate as a conjecture, proved the failure in d=1, and gave the semiclassical link to the transport equation via Proposition 5.1."},{"cited_title":"Bennett, N","cited_arxiv_id":null,"evidence_quote":"Conjectured the kinetic transport analogue in dual form, which Corollary 1.9 answers."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Earlier abstract orthonormal Strichartz result in a semiclassical setting, used as the comparison baseline showing the missing endpoint."}],"review_version":1}