{"id":"f0474136-980b-4aa1-b1fe-135087479100","arxiv_id":"2508.02663","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":4.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":2,"one_line_summary":"BTZ black hole microstates under collective-field boundary conditions are labeled by Young diagrams, and the logarithmic correction to their entropy is -1/2, one-loop exact and identical for two boundary Hamiltonians.","lead":"This paper derives the canonical partition function and logarithmic entropy correction for BTZ black holes in AdS3 gravity with collective-field boundary conditions. It shows the microstates are counted by Young diagrams and that the one-loop correction coefficient is -1/2 for two different boundary Hamiltonians.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The -1/2 log correction rests on an unproven identification of (6.5) with chiral U(N) Yang-Mills; the O(1/N) term dropped in (6.6) and the cutoff prescription could alter the one-loop coefficient.","rationale":"I agree with the reader that the ColFT-to-Yang-Mills mapping is the softest point of the argument. The rest of the paper is internally consistent: the microcanonical degeneracy count via Hardy-Ramanujan reproduces the Bekenstein-Hawking entropy, and the relativistic partition function in (6.20)-(6.22) gives the same -1/2 coefficient by a direct and transparent computation. The YM identification is the only step that imports an external genus expansion, and the paper itself flags it with 'closely resembles', an unevaluated sum over higher genera, and unidentified extra saddles. I do not see an internal contradiction or a clear numerical error; the issue is an unproven equality at subleading order and a possible sensitivity to O(1/N) corrections in the rank/area identification. These considerations support keeping the reader's CONDITIONAL verdict: the central claim may well be correct, but it should be secured by an exact derivation or a numerical test before being accepted as established.","tokens_in":22191,"tokens_out":29025,"duration_ms":334408,"concrete_test":"Evaluate (6.5) exactly at fixed N=|n| (e.g., N=100 and 200) over all representations with at most N rows, keeping the full (2N+1)|R| term and no large-N cutoff, for areas A=2*pi/(sqrt(3)*N) and neighboring values; then test whether ln Z_exact is reproduced by N^2*F0(A) + F1(A) + N^{-2}*F2(A) with F1(A)=-log eta(e^{-A/2}) to subleading accuracy. If the residual contains a term proportional to log A with coefficient different from 1/2, the one-loop claim fails; if the residual is O(1) as N goes to infinity with no log A, the identification is confirmed.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The central claim—coefficient -1/2, one-loop exact—is inherited from the genus expansion of 2D U(N) Yang-Mills on a torus, applied to the sum in (6.5). But the identification of that sum with the chiral U(N) YM partition function is not derived; Section 6.1 introduces (6.6) as 'this expression closely resembles' and cites [44-51] rather than proving the equality. Concretely, (6.5) contains the exact term (2|n±|±1)|R|, while the YM form (6.6) is obtained only after dropping the ±1 using |n±|≫1; this is an O(1/N) change in the effective rank and area of the gauge theory. The row sum is also regulated by a cutoff i<|n±| and then the limit is taken, which is not automatically the same as the U(N) representation sum at fixed large N. If the exact sum differs from the chiral-YM partition function by a representation-dependent measure or by an N-dependent renormalization of the area, the F1 term -log eta(Q) and its 1/2 log A coefficient need not survive. The relativistic-fermion calculation in Section 6.2 is independent and does give -1/2, but the 'universal and one-loop exact' statement for the ColFT Hamiltonian depends entirely on the unproven YM mapping.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"This paper studies asymptotically AdS3 gravity under non-standard boundary conditions in which the chemical potentials are determined dynamically by a chosen boundary Hamiltonian. Taking the boundary Hamiltonian to be the collective-field-theory (ColFT) Hamiltonian, the authors show that BTZ black holes correspond to constant-density, constant-velocity fluid configurations and fix the proportionality constants C± so that the bulk metric takes the standard BTZ form. Quantizing via bosonization of relativistic fermions, they construct a Hilbert space labeled by a U(1) charge n± and identify black-hole microstates with particle-hole excitations organized into Young diagrams; Hardy-Ramanujan counting of these diagrams reproduces the Bekenstein-Hawking entropy at leading order. The new results concern Euclidean canonical partition functions. For the ColFT Hamiltonian, the sector partition functions Z± are written as sums over Young diagrams, asserted to resemble the partition function of chiral U(N) Yang-Mills theory on a torus with N±=|n±|, and expanded in a genus expansion; the paper claims the leading free energy receives contributions from all genera (with the higher-genus part left as an unevaluated sum), while the logarithmic correction comes solely from the genus-one term and is one-loop exact with coefficient −1/2.","tokens_in":22485,"tokens_out":36383,"duration_ms":362519,"significance":"The paper contains several solid and useful ingredients. The microcanonical counting of Section 5.4 is explicit and checkable: the map |R±|=(c/2)(Ml±J) together with the Hardy-Ramanujan estimate gives exactly S=π²l²/(2G)(1/β+ + 1/β−), a nontrivial consistency check between the Young-diagram Hilbert space and the Bekenstein-Hawking formula. The relativistic-channel computation in Section 6.2 is the strongest part: the identity Z±=Π_{k≥1}(1−e^{−µ±k})^{−1} is exact, and the modular transform of the Dedekind eta function delivers the coefficient −1/2 without uncontrolled approximation. This gives a concrete, falsifiable prediction distinguishing these Kac-Moody-type boundary conditions from the Cardy/Carlip −3/2 scenario. If the identification of the ColFT representation sum with chiral U(N) Yang-Mills theory can be made exact, the paper would establish a new bridge between BTZ microstates and the 2D Yang-Mills/topological-string genus expansion. As it stands, the ColFT channel is a plausible conjecture that is consistent with, but not independently established by, the rigorous relativistic-channel result.","major_comments":[{"comment":"The claim that the ColFT-channel logarithmic correction is one-loop exact with coefficient −1/2 rests on identifying the sum (6.5) with the partition function of chiral U(N) Yang-Mills theory on a torus, but this identification is asserted rather than derived. The passage from (6.5) to (6.6) replaces the coefficient (2|n±|±1) by 2|n±|, thereby dropping a term ±|R±| of relative order 1/|n±| in the exponent, and imposes a cutoff i<|n±| on the row sum that is not present in the Hilbert-space trace. The dropped term is not negligible: at the saddle point of the representation sum one has |R±| ~ π²/(6ñ²) and N±Ã±=2π/√3, so the omitted term contributes an O(1) amount to the exponent. In addition, the matching of rank and area is not stated precisely (the coefficient of |R±| in (6.6) differs by a factor of 2 from the standard U(N) quadratic-Casimir expression, which would require a compensating rescaling of ñ), and the genus expansion quoted from Refs. [52-54] is applied in the regime ñ→0 with N±Ã± fixed rather than in the fixed-area large-N limit in which it is normally derived. Since the asserted one-loop exactness of the −1/2 coefficient is precisely what is at stake, the authors should either prove that (6.6) is exactly the chiral U(N) amplitude (with the correct rank, area, and row cutoff) or provide a controlled estimate showing that the O(1/N) modifications and the cutoff do not affect the coefficient of log ñ in F1.","section":"Section 6.1, Eqs. (6.5)-(6.6), footnote 4"},{"comment":"For the ColFT Hamiltonian the leading-order free energy is not actually computed. Equation (6.14) expresses F± as the genus-one term π²/(3ñ) plus an infinite sum over genus g≥2 whose coefficients c_g are not known and whose convergence is not established, as the authors themselves acknowledge. The claim that the genus-one part reproduces the BTZ free energy while the remainder represents additional saddle points is therefore a conjecture: if the remainder is non-zero, the leading-order free energy (and hence the leading entropy) would not match the Bekenstein-Hawking value. The abstract's statement that the leading entropy term receives contributions from all genera is a structural statement consistent with Eq. (6.11), but it should be accompanied by an explicit statement that the quantitative value of the leading term, and in particular the BTZ matching, remains an open problem for the ColFT case. This caveat does not affect the logarithmic coefficient itself, since the higher-genus terms carry no logarithm.","section":"Section 6.1, Eqs. (6.13)-(6.15)"},{"comment":"The canonical partition function is posited as Z=Tr exp(−β+H+−β−H−) over the bosonized boundary Hilbert space, and its equivalence to the Euclidean bulk Chern-Simons path integral under the ColFT boundary conditions is not derived. The paper shows only the classical equivalence between −βM+βΩJ and −β+H+−β−H−; the quantum measure, including the representation sum and the background subtraction that fixes n± through Eq. (6.2), is assumed. Because logarithmic corrections are sensitive to measure factors (a missing β±-dependent normalization of the trace could change the log coefficient), the paper should state explicitly whether (6.1) is intended as a definition of the quantum boundary theory or as a derived statement, and in the latter case indicate the derivation or its limitations.","section":"Section 6.1, Eqs. (6.1)-(6.3)"}],"minor_comments":[{"comment":"There is a duplicated phrase 'In in our analysis', and the acknowledgments sentence should read 'We thank Nabamita Banerjee and Ranveer Singh for useful discussions'.","section":"Section 5.3 and Acknowledgments"},{"comment":"The notation l±i is confusing because l_i already carries the row index i; a notation such as l_i^{(±)} would make the two chiral sectors easier to follow.","section":"Equation (6.4)"},{"comment":"The factor e^{−t/24} uses an undefined variable t and should read e^{−µ/24}.","section":"Appendix B, last displayed formula"},{"comment":"The equivalence of the intermediate form (1/2) ln(β+β−/(c²l²)) with the final form −(1/2) ln(9l⁴/(4G²β²(1−l²Ω²))) uses β+β−=β²(1−l²Ω²) and c=3l/(2G); these substitutions should be stated explicitly in the text.","section":"Equations (6.16) and (6.24)"},{"comment":"The Hardy-Ramanujan asymptotics includes a prefactor 1/(4√3 n) that would contribute −(1/2) log n to the microcanonical entropy; since this has the same coefficient as the canonical result, a remark on the consistency (or a definition of the precision at which the microcanonical match is claimed) would be helpful.","section":"Section 5.4"},{"comment":"The statement that the degeneracy of these states 'exactly reproduces the classical Bekenstein-Hawking entropy' is too strong in view of the leading-order Hardy-Ramanujan estimate; 'reproduces at leading order' would be more accurate.","section":"Section 7"}],"recommendation":"major_revision","confidential_remarks":"The relativistic-channel result (Section 6.2) is solid and could stand on its own as a short paper. The value added by the ColFT channel is the attempted universality claim, but Section 6.1 currently supports only a resemblance, not an equality; the abstract and conclusion should either be brought in line with a clearly labeled conjecture or the identification should be proven. The authors are candid in the text about the unevaluated higher-genus sum, and that candor should be preserved in the abstract. No issues with novelty disclosure: the 'Note added' and the reference to [21] make the incremental character of the work clear. The paper's fit with hep-th is appropriate."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"You should know two things about this paper. First, the half that is rigorous is genuinely nice: the free-fermion boundary Hamiltonian gives a Young-diagram counting of BTZ microstates that reproduces the Bekenstein-Hawking entropy, and the canonical partition function is exactly the generating function for integer partitions, from which the -1/2 logarithmic correction follows cleanly via the modular transformation of the Dedekind eta function. Second, the more novel half—the ColFT Hamiltonian—is a lot softer. The partition function is argued to resemble chiral U(N) Yang-Mills on a torus, and on that resemblance the paper hangs both the leading free energy and the claim that the -1/2 correction is one-loop exact. But the resemblance is never upgraded to a derivation.\n\nWhat is actually new: the canonical partition function for the ColFT case and its interpretation through the 2D Yang-Mills genus expansion. The Young-diagram microstate counting was already in their earlier paper [21], and the -1/2 log correction for a single U(1) Kac-Moody boson is a known result. They cite [28] for bosonization but never compare their coefficient with that paper's horizon-fluff result, which is a missed connection.\n\nSoft spots, in order. (1) Equation (6.5) is converted to the Yang-Mills form (6.6) by dropping the ±1 in (2|n|±1), an O(1/N) change. In the classical limit A ~ 1/N, that shift affects the free energy at O(1); it probably does not change the coefficient of log A, but the one-loop-exact claim is only as good as the dropped term plus the cutoff on the row sum, which is consistent with U(N) but not justified from the bulk. (2) The leading free energy in the ColFT case is an unevaluated infinite series of higher-genus coefficients; the paper is honest about this, but it means the 'leading' term is not actually computed. (3) The YM identification is stated as 'closely resembles' with references, not derived from the Chern-Simons path integral. If there is a representation-dependent measure missing, the one-loop coefficient could change.\n\nNone of this breaks the relativistic calculation, which stands on its own. But the universality claim is only as strong as the ColFT step, and that step is conditional.\n\nThis paper is for people working on AdS3 microstates, non-standard boundary conditions, and log corrections. It deserves a serious referee, though the referee should push for a derivation of the YM equivalence and a comparison with [28]. I would not cite it in my own work before those points are cleared up.\n\nRecommendation: send to peer review, flagging the YM identification as the main issue.","headline":"The relativistic-fermion side is rigorous and clean; the ColFT claim of a universal -1/2 log correction rests on an asserted Yang-Mills identification that needs a real derivation.","tokens_in":23053,"tokens_out":14134,"would_cite":false,"duration_ms":124470,"reading_group":"maybe","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":["81T40","83C57","81T13"],"pacs":["04.70.Dy","04.60.-m","11.25.Tq","11.15.-q"],"model":"deepseek-v4-flash","headline":"This paper claims that the logarithmic correction to BTZ black hole entropy is exactly -1/2 under two different boundary Hamiltonians, and that this correction is one-loop exact.","keywords":["BTZ black hole","AdS3/CFT correspondence","bosonization","collective field theory","black hole microstates","logarithmic entropy correction","Young diagrams","2D Yang-Mills theory"],"falsifier":"Compute the finite-$N$ version of the sum in (6.5) numerically for large but finite $|n_\\pm|$ and extract the coefficient of $\\log\\beta$; a value different from $-1/2$ would show the chiral Yang–Mills identification receives subleading corrections. Alternatively, evaluate the next term in the small-area expansion of the ColFT free energy directly from the Young-diagram sum and compare it with the genus-expansion prediction, since a mismatch would indicate the all-genus leading term is not controlled by the two-dimensional Yang–Mills sector.","tokens_in":21971,"feed_emoji":"🕳️","tokens_out":8686,"duration_ms":72878,"temperature":0.7,"pith_summary":"This paper tries to show that the microstates of a BTZ black hole can be counted by a one-dimensional fermionic fluid on the boundary of AdS$_3$, and that the bulk thermodynamics follows from the fluid's quantum mechanics. Quantizing the collective-field description through bosonization produces states labeled by Young diagrams whose degeneracy reproduces the Bekenstein–Hawking entropy. The same construction yields a canonical partition function that, for the collective-field Hamiltonian, takes the form of a chiral $U(N)$ Yang–Mills theory on a torus with $N \\sim 1/(\\beta G)$. From that partition function the paper derives a logarithmic correction to the entropy with coefficient $-1/2$ that comes entirely from the genus-one sector and is unchanged when the boundary Hamiltonian is replaced with a relativistic free-fermion Hamiltonian. A sympathetic reader would care because the paper gives a concrete boundary model in which both the leading entropy and the one-loop quantum correction can be computed explicitly.","feed_headline":"Log correction to BTZ entropy is -1/2, one-loop exact","feed_subtitle":"Bosonized boundary fluid counts BTZ microstates via Young diagrams and ties the partition function to 2D Yang-Mills.","key_machinery":"The load-bearing machinery is the bosonization identity $:\\psi^\\dagger\\psi:=\\sqrt{c}\\,\\tilde p$, which converts the boundary Kac–Moody current into a fermion bilinear and builds the Hilbert space from particle–hole excitations above an $n$-particle ground state $|n\\rangle$. These excitations are labeled by Young diagrams, with the number of boxes encoding the mass and angular momentum of the black hole. For the canonical ensemble, the partition-function sum in (6.5) is identified with the chiral $U(N)$ Yang–Mills partition function on a torus, with $N=|n|$ and area $\\tilde A_\\pm=2\\beta_\\pm/(cl)$, whose genus expansion supplies the leading and one-loop contributions to the free energy.","core_discovery":"The central discovery is that the Euclidean canonical partition function of the BTZ black hole, computed with Kac–Moody boundary conditions determined by a collective-field boundary Hamiltonian, factorizes into two chiral sectors whose sums over Young diagrams reproduce the partition function of chiral $U(N)$ Yang–Mills theory on a torus, with rank $N\\sim c/(\\beta l)\\sim 1/(\\beta G)$. Using the known genus expansion of two-dimensional Yang–Mills, the paper shows that the leading free energy receives contributions of the same order from all genera, while the subleading logarithmic term comes only from the genus-one sector and has coefficient $-1/2$. For a second boundary Hamiltonian describing relativistic fermions, the partition function reduces to the generating function for integer partitions, and the same $-1/2$ logarithmic correction appears; the paper interprets this agreement as universality of the one-loop correction across boundary Hamiltonians. In the same framework, microstates of the black hole are particle–hole excitations of the fermionic system, labeled by Young diagrams, whose degeneracy matches the Bekenstein–Hawking entropy.","pith_inferences":["Editorial extension: the chiral $U(N)$ Yang–Mills form suggests that the boundary partition function may admit a nonperturbative resummation using modular or topological-string techniques, going beyond the genus-by-genus treatment in the paper.","Editorial extension: because the $-1/2$ coefficient comes solely from the genus-one sector, a natural test is to compute the logarithmic correction for a third boundary Hamiltonian, such as an interacting fermionic Hamiltonian, and check whether the coefficient remains $-1/2$.","Editorial extension: the claim that all genera contribute at the same order at leading entropy implies the classical limit is not a single saddle; identifying the missing saddle points could connect these microstates to known families of Euclidean geometries in AdS$_3$."],"forward_implications":["If the identification with chiral $U(N)$ Yang–Mills is exact, the full partition function of the BTZ black hole under these boundary conditions is known to arbitrary order in the genus expansion.","The degeneracy of Young diagrams with fixed box number reproduces the Bekenstein–Hawking entropy, giving a microscopic count that is tied to a concrete fermionic Hilbert space rather than to an asymptotic Cardy-type formula.","The logarithmic correction has coefficient $-1/2$ and is one-loop exact, and it is the same for the collective-field and relativistic-fermion Hamiltonians, so it is a robust prediction of these Kac–Moody boundary conditions.","For the collective-field Hamiltonian the leading free energy includes contributions from all genera, which the paper interprets as additional saddle points in the bulk path integral that have not yet been identified.","For the relativistic-fermion Hamiltonian the free energy matches the standard BTZ free energy without extra saddles, isolating the effect of the boundary Hamiltonian on the classical thermodynamics."],"supporting_citations":[{"why":"Introduces the collective-field boundary conditions and the fluid-dynamics picture on which the paper's quantization is built.","marker":"[18]"},{"why":"Supplies the collective field theory Hamiltonian used as the boundary Hamiltonian.","marker":"[19,20]"},{"why":"The authors' earlier construction of BTZ microstates that the present paper extends to the full canonical partition function.","marker":"[21]"},{"why":"Defines the higher bosonic mode operators B^K_n used to express the boundary fields after bosonization.","marker":"[31]"},{"why":"The two-dimensional Yang–Mills partition function to which the ColFT partition function is compared.","marker":"[44–51]"},{"why":"Provides the genus expansion and small-area asymptotics of the torus Yang–Mills free energy used for the leading and one-loop terms.","marker":"[53]"},{"why":"A similar bosonized construction of near-horizon microstates with logarithmic corrections that motivates the Hilbert space.","marker":"[28]"},{"why":"The single-CFT benchmark giving log coefficient -3/2 that the paper's -1/2 result is measured against in the discussion of logarithmic corrections.","marker":"[42]"}],"fun_headline_variants":["BTZ entropy log correction is -1/2, one-loop universal","One-loop -1/2 log entropy holds across BTZ Hamiltonians","Black hole entropy: -1/2 log term is one-loop exact","BTZ microstates counted: log correction -1/2 from one loop","Universal -1/2 log correction to BTZ entropy via bosonization"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The calculation rests on identifying the boundary trace over bosonized fermion states with the Euclidean bulk path integral, and on treating the sum in (6.5) as exactly the chiral $U(N)$ Yang–Mills partition function; if either identification fails at subleading order, the coefficient of the logarithmic correction would change.","fun_headline_variants_meta":{"raw":{"variants":["BTZ entropy log correction is -1/2, one-loop universal","One-loop -1/2 log entropy holds across BTZ Hamiltonians","Black hole entropy: -1/2 log term is one-loop exact","BTZ microstates counted: log correction -1/2 from one loop","Universal -1/2 log correction to BTZ entropy via bosonization"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.001032,"raw_usage":{"total_tokens":4370,"prompt_tokens":990,"completion_tokens":3380,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":606,"completion_tokens_details":{"reasoning_tokens":3264}},"tokens_in":606,"tokens_out":3380,"duration_ms":22639,"temperature":1.0,"reasoning_tokens":3264,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-15T17:38:52.446532+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Compute the finite-$N$ version of the sum in (6.5) numerically for large but finite $|n_\\pm|$ and extract the coefficient of $\\log\\beta$; a value different from $-1/2$ would show the chiral Yang–Mills identification receives subleading corrections. Alternatively, evaluate the next term in the small-area expansion of the ColFT free energy directly from the Young-diagram sum and compare it with the genus-expansion prediction, since a mismatch would indicate the all-genus leading term is not controlled by the two-dimensional Yang–Mills sector.","supporting_citations":[{"cited_title":"Higher Spin Gravity in $AdS_3$ and Folds on Fermi Surface","cited_arxiv_id":"2302.08471","evidence_quote":"Introduces the collective-field boundary conditions and the fluid-dynamics picture on which the paper's quantization is built."},{"cited_title":"Bosonisation and BTZ Black Hole Microstates","cited_arxiv_id":"2502.19322","evidence_quote":"The authors' earlier construction of BTZ microstates that the present paper extends to the full canonical partition function."},{"cited_title":"The String partition function for QCD on the torus,","cited_arxiv_id":null,"evidence_quote":"Provides the genus expansion and small-area asymptotics of the torus Yang–Mills free energy used for the leading and one-loop terms."}],"review_version":2}