{"id":"b34c47d7-d364-42af-9e8b-3545b1eba133","arxiv_id":"2512.23528","paper_version":2,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":0,"one_line_summary":"For freely independent selfadjoint x and y with y free Poisson, the absolutely continuous part of the Brown measure of x + i y has density expressible through the inverse of an explicitly constructed map h.","lead":"This paper derives an explicit formula for the density of the Brown measure of x + i y, where x is an arbitrary selfadjoint element and y is free Poisson. The formula uses an invertible map built from a subordination left inverse, with the full case x symmetric Bernoulli worked out.","discovery_kind":"new_method","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Proposition 5.2's determinant proof contains an apparent algebraic error in its row reduction, leaving the strict positivity of the Jacobian—and hence the reparametrization/density formula—unproved.","rationale":"I read the paper in good faith and followed the main line: the Brown measure density is derived from the Laplacian of L_{x+iy}, using subordination to express derivatives in terms of h^{-1}. The formula in Theorem 6.4 is internally consistent, and the abstract's version with t in the denominator matches the proof; the typo in Theorem 1.8 (s instead of t) is real but non-central because Theorem 6.4 restates the correct formula. The reader's weakest_assumption isolates Proposition 5.2, and my independent check confirms that this is indeed the most load-bearing step: the strict positivity of the 3×3 Jacobian is what licenses the inverse function theorem, the differentiability of h^{-1}, and the real-analytic extension used to justify differentiating under the limit. That proof is sketched with a row reduction that appears algebraically incorrect unless an unstated relation at δ0 holds. The paper contains a detailed worked example with a numerical eigenvalue simulation, which is valuable supporting evidence for the specific Bernoulli case, but it does not establish the general determinant positivity. The open atom question for p<1 and the conditional Assumption 1.6 for p≤1 further temper the scope, but the determinant issue is the clearest correctness risk. Since this is a genuine proof gap rather than a demonstrated counterexample, the appropriate verdict remains conditional, matching the reader's assessment.","tokens_in":24310,"tokens_out":4475,"duration_ms":40892,"concrete_test":"Recompute det JĤ(α,β,δ0) symbolically for the symmetric Bernoulli example in §1.4 using the explicit formulas for H12, H11, and δ0, and verify positivity on a fine grid covering D. Then repeat for a second measure, e.g. µ_x uniform on two asymmetric atoms or a two-point measure with unequal weights, with p>1, again checking positivity numerically on a grid. Independently, rederive the determinant reduction of Proposition 5.2 without the contested row operation; if the determinant cannot be reduced to a manifestly positive expression, the proof needs repair. This check settles whether the Jacobian obstruction is real.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The central reparametrization h:D→M and the real-analyticity of L_{x+iy} (and therefore the density formula) hinge on Proposition 5.2's claim that det JĤ(α,β,δ0)>0 for α+iβ∈D. The proof's key row reduction is not justified. Starting from the displayed 3×3 matrix, the text performs r2 := r2 − (β/δ)r3 and claims that row 2 becomes [−∂D/∂α, −∂D/∂β, −∂D/∂δ]/D². Direct computation gives, for example, the (2,1) entry as [−(1−βT)+βT/δ]·∂D/∂α /D², not −∂D/∂α/D². The extra βT/δ term does not vanish at δ0(λ); at that point only T/D=1/p is known. A similar discrepancy affects the other entries. Consequently, the subsequent elimination of all ∂D terms and the final 'Expanding along the second column … using Cauchy-Schwarz' step are unsupported. If the determinant can vanish for some admissible µ_x, the inverse-function-theorem step in Lemma 6.2/Proposition 6.3 fails, and the density formula is not established. This is a proof gap, not a refutation—the worked Bernoulli example may still satisfy positivity—but the general claim rests on an unverified computation.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper studies the Brown measure of a=i x? of the form x+iy, where x is an arbitrary selfadjoint element and y is a free Poisson element of parameter p, in a tracial W*-probability space. The authors use the matrix-valued Hermitization of a and operator-valued subordination to construct a left inverse H of the subordination function. They define an open set D and a map h:D→M by solving H_{11}(λ,iδ0(λ))=0 and setting h(λ)=H_{12}(λ,iδ0(λ)). The main theorem, Theorem 1.8/6.4, asserts that the Brown measure is supported in cl(M) and that, for s+it∈M, its absolutely continuous density is f=(1/(4π))[(2/t)(∂α/∂s+∂β/∂t)-2/t-2β/t²], where α+iβ=h^{-1}(s+it). The proof proceeds by showing that the logarithmic potential extends real-analytically across M, computing its first derivatives via subordination, and taking the Laplacian. A fully explicit Bernoulli example is worked out, with formulas for D, h, the support curve, and the density, and the support is identified with the spectrum.","tokens_in":24664,"tokens_out":15352,"duration_ms":127859,"significance":"If the main theorem is correct, this is a substantial contribution: it gives a constructive, checkable formula for the absolutely continuous part of the Brown measure for a whole class of non-normal free elements, going beyond the semicircular and elliptic cases. The methodology is not circular: h is constructed explicitly from the left inverse H, no parameters are fitted, and the Bernoulli example is closed-form and agrees with numerical random-matrix simulations. The paper is clearly organized and the subordination framework is natural. The main claims are falsifiable and the explicit example is a valuable sanity check. The main caveat is that one load-bearing Jacobian-positivity proof is only sketched, and the theorem statement contains a typo in the central formula.","major_comments":[{"comment":"The strict positivity of det J bH at δ0 is the load-bearing point for the inverse-function step in Lemma 6.2 and hence for the real-analyticity of L and the density formula. The proof's final step is only asserted: after the row and column reductions it says 'Expanding along the second column the result follows...'. I checked the row reduction r2 := r2 - (β/δ)r3; it is actually correct, giving row 2 = [-D_α, -D_β, -D_δ]/D². The subsequent column operation c2 := c2 - (β/δ)c3 is also correct. However, the final determinant is not displayed. Writing Q=(α-t)²+β²+δ² and ν=dµ_x(t)/Q², the reduction yields det = 8Tδ²(∫1 dν · ∫(α-t)² dν - (∫(α-t) dν)²)/D⁴, which is nonnegative by Cauchy-Schwarz and positive in the nondegenerate case. This computation should be included explicitly; as written, the proof of a central analytic fact is an unverified calculation.","section":"§5.1, Proposition 5.2"},{"comment":"The displayed density in the the statement of Theorem 1.8 is f = (1/(4π))[(2/t)(∂α/∂s+∂β/∂t) - 2/s - 2β/s²]. The denominators should be t, not s: the abstract and Theorem 6.4 have -2/t - 2β/t². As stated, readers using Eq. (1.9) will get a wrong density. This must be corrected in the final version.","section":"§1.3, Theorem 1.8, Eq. (1.9)"},{"comment":"Neither theorem lists Assumption 1.6 (spec(x)⊆cl(D)) as a hypothesis, but the proof that the Brown measure has no mass outside cl(M) is Proposition 6.5, which explicitly invokes Assumption 1.6. The abstract acknowledges that conditions on x are needed for some p, but the theorem statements as written appear unconditional. The theorems should either state Assumption 1.6 explicitly or be formulated as conditional statements, with the Section 4 sufficient conditions stated separately.","section":"§1.3 and §6, Theorems 1.8 and 6.4"}],"minor_comments":[{"comment":"In Eq. (6.3) and the displayed formula for L_{x+iy}(z,0), the integrand contains 1/(1+t) when the integration variable is u. It should be 1/(1+u).","section":"§6, Lemma 6.1 and Lemma 6.2"},{"comment":"The displayed matrices in the proof omit parentheses in expressions like -δ ∂T/∂α D - T ∂D/∂α; this makes the row operations hard to follow. Please write these as -δ(∂T/∂α · D - T · ∂D/∂α) etc.","section":"§5.1, Proposition 5.2"},{"comment":"There are numerous typos: 'operator velued' (§1.5), 'F ree Poisson' (§2.3), 'differmorphism' (§5.3), 'measre' (§1.5), 'exits' (§6, Lemma 6.1). A careful proofreading pass is needed.","section":"Throughout"},{"comment":"The sentence 'We will clarify this in later section' looks like a leftover from an earlier draft and should be removed or replaced by a reference to Proposition 4.3/4.5.","section":"§4, Remark 4.4"}],"recommendation":"major_revision","confidential_remarks":"The stress-test concern about Proposition 5.2 does not land as an algebraic error: the row reduction is valid, and the final positivity follows from a short Cauchy-Schwarz computation. The proof nevertheless needs to be written out because it is load-bearing. With that addition plus the typo and hypothesis-statement fixes, the paper should be acceptable. I saw no circularity or parameter-fitting issue."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Here's my take. The paper does something new: it gives an explicit absolutely continuous density for the Brown measure of x+iy when y is free Poisson, with a reparametrization h constructed via the left inverse H of the subordination function. That's not in the existing literature — the semicircular and elliptic cases were known. The worked Bernoulli/p=1 example is fully explicit and matches the matrix model. The construction of h from H is a genuinely useful device; it plausibly generalizes to other y.\n\nThe main formula is derived, not fitted. Subordination and Hermitization lead to the density in terms of h^{-1}. No parameters are fit. Good.\n\nThe soft spots: First, the abstract and Theorem 1.8 contain a typo: the density formula has -2/s - 2β/s^2, which should be -2/t - 2β/t^2. Theorem 6.4 has the correct version, so it's a copy-paste error, but in a theorem statement it matters.\n\nSecond, Proposition 5.2, the positivity of the Jacobian, is the load-bearing step, and as printed the proof has a real problem. The stress-test note is right: the row operation r2 := r2 - (β/δ)r3 does not produce the simplified second row from the displayed matrix. But the issue is not the method — it's a sign error in the displayed third row. Differentiating bH3 = δ - pδT/D with respect to α gives -δ T_α D + δ T D_α over D^2, not -δ T_α D - T D_α over D^2. With the plus sign restored, the same row operation works exactly as the text claims, and the subsequent Cauchy-Schwarz argument is plausible. So the gap is a typo, not a fatal flaw — but it needs to be fixed because no one can verify the determinant from the printed version. I'd want to see the corrected calculation or a computer algebra check of det J > 0 before relying on Proposition 5.2.\n\nThird, Assumption 1.6 is conditional for p≤1, and the atom question for p<1 is left open. They're honest about this, and the paper's main result is stated with the assumption. That's fine, but it limits the scope: for p≤1 (except when the sufficient conditions hold) the density formula is conditional.\n\nFourth, the numerical check is a heuristic; it matches the boundary and density shape, but it doesn't verify the formula. That's okay — the derivation is the argument — but I wouldn't call it confirmation.\n\nThe citation pattern looks fine; they cite the relevant subordination and Brown-measure literature, and the self-citations are to prior results in the same line. I don't see a circularity problem: the density is computed from definitions, not fit to the example.\n\nBottom line: this is a solid, new computation with a clear main theorem, a beautiful worked example, and a proof that's fixable but has typos in a critical place. It deserves peer review. A referee should be asked to verify Proposition 5.2 with the corrected signs and to check whether the positivity can be proved directly. I would not desk-reject. I'd also tell the authors to fix the s/t typo in the abstract and Theorem 1.8 before resubmission.","headline":"A genuinely new Brown-measure computation for x+iy with y free Poisson; the main theorem is right in substance, but Proposition 5.2 has sign typos that make the printed proof hard to trust, and the abstract/Theorem 1.8 have a variable typo.","tokens_in":25144,"tokens_out":6984,"would_cite":true,"duration_ms":52045,"reading_group":"maybe","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":["46L54","47A10"],"pacs":[],"model":"deepseek-v4-flash","headline":"A subordination left inverse yields an explicit density formula for the Brown measure of x + i y with y free Poisson.","keywords":["Brown measure","free Poisson","subordination","free probability","operator-valued Cauchy transform","non-Hermitian random matrices","Brown spectral measure"],"falsifier":"For a concrete µ_x (say x with symmetric Bernoulli law), compute det J Ĥ(α,β,δ_0(α+iβ)) from Proposition 5.2 at a dense grid of points of D; finding any point where it is zero or negative would falsify the claim. A softer check: verify numerically that the density formula integrates to the total Brown mass over M.","tokens_in":24218,"feed_emoji":"🧮","tokens_out":4666,"duration_ms":38308,"temperature":0.7,"pith_summary":"The paper establishes that, when y is a free Poisson element freely independent of a selfadjoint x, the Brown measure of x+iy has its support inside the closure of a region M, and on M its density is given by an explicit formula in terms of the inverse of a change of variables h. This matters because explicit Brown-measure computations are rare: most known cases rely on special structure, whereas here the formula is systematically derived from the matrix-valued subordination function of the hermitization. The key move is to use the explicit left inverse H of the subordination function to construct h, so the density becomes computable from the derivatives of h^{-1}. A fully worked example (x with symmetric Bernoulli law, p=1) produces a concrete algebraic support and density matching random-matrix simulations.","feed_headline":"Explicit density for Brown measure of x + i·free Poisson","feed_subtitle":"A subordination left inverse turns the Brown-measure density into a formula in terms of h^{-1}; Bernoulli case worked out.","key_machinery":"The engine is the matrix-valued subordination function Ω of the hermitization of x+iy, together with its explicit left inverse H(B)=B+p(J−G_X(B))^{-1} obtained from the rational R-transform of the free Poisson element. H is used to define the domain D (points λ where a certain unimodular-type limit is negative), the positive function δ_0(λ) satisfying H_11(λ, iδ_0)=0, and the reparametrization h(λ)=H_12(λ, iδ_0(λ)). The identity H(Ω(B))=B and the strict positivity of a 3x3 Jacobian (Proposition 5.2) make h a diffeomorphism and allow the logarithmic potential L_{x+iy} to be extended real-analytically, from which the density is read off by applying the Laplacian.","core_discovery":"The central result is Theorem 1.8 (restated as Theorem 6.4): for s+it in the open set M=h(D), the absolutely continuous density of the Brown measure of x+iy is f(s,t)=1/(4π)[(2/t)(∂α/∂s+∂β/∂t)−2/t−2β/t^2], where α+iβ=h^{-1}(s+it). Here D and M are open subsets of C built from the left inverse H(B)=B+p(J−G_X(B))^{-1} of the subordination function Ω of the hermitization of x+iy, and h(λ) is the off-diagonal entry of H at the unique δ_0(λ)>0 where the diagonal entry vanishes. The support of the Brown measure is contained in cl(M), and under Assumption 1.6 (spec(x)⊆cl(D), automatic when p>1) the region outside cl(M) carries no mass. In the Bernoulli example the support coincides with the spectru","pith_inferences":["A direct numerical test of the density formula would be to compute h^{-1} for a grid in M and compare the resulting f with eigenvalue histograms of large random-matrix models; the paper's own Bernoulli figure does this for one case, but other µ_x are untested.","The identity between the Brown-measure support and the spectrum, verified in the Bernoulli example, is suggested as a general phenomenon; the paper only proves support containment, so checking spectrum-vs-support agreement for other µ_x would sharpen the statement.","The method's reliance on an explicit left inverse H suggests that other distributions of y with a tractable rational R-transform (for instance, certain elliptic or compound free Poisson laws) may admit analogous density formulae."],"forward_implications":["If the main theorem is correct, the Brown measure of x+iy is fully described by an explicit, checkable formula whenever one can invert h — no limiting random-matrix heuristics needed.","For p>1, the computation is unconditional in the sense that Assumption 1.6 is automatic; for p≤1 the supplementary conditions on µ_x are explicit and testable.","The Bernoulli example shows the method produces an algebraic boundary for the support and a closed-form density whose singular behavior at ±1 is visible in simulations.","The same subordination-left-inverse strategy is proposed as a general methodology that may extend to other distributions of y, including the semicircular case."],"fun_headline_variants":["Free Poisson makes Brown measure density explicit","Subordination inverse yields explicit Brown measure density","Explicit Brown-measure density via subordination inverse","Brown measure density from a subordination trick"],"cache_read_input_tokens":2304,"weakest_assumption_plain":"The whole construction rests on the unproved assertion in Proposition 5.2 that the 3x3 Jacobian of the reparametrized map is strictly positive inside D; if that determinant can vanish for some distribution of x, the diffeomorphism h and the density formula collapse.","fun_headline_variants_meta":{"raw":{"variants":["Free Poisson makes Brown measure density explicit","Subordination inverse yields explicit Brown measure density","Explicit Brown-measure density via subordination inverse","Brown measure density from a subordination trick"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.001013,"raw_usage":{"total_tokens":4222,"prompt_tokens":955,"completion_tokens":3267,"prompt_tokens_details":{"cached_tokens":256},"prompt_cache_hit_tokens":256,"prompt_cache_miss_tokens":699,"completion_tokens_details":{"reasoning_tokens":3211}},"tokens_in":699,"tokens_out":3267,"duration_ms":22462,"temperature":1.0,"reasoning_tokens":3211,"cache_read_input_tokens":256,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-03T13:38:08.430991+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"For a concrete µ_x (say x with symmetric Bernoulli law), compute det J Ĥ(α,β,δ_0(α+iβ)) from Proposition 5.2 at a dense grid of points of D; finding any point where it is zero or negative would falsify the claim. A softer check: verify numerically that the density formula integrates to the total Brown mass over M.","supporting_citations":[],"review_version":1}