{"id":"d327651c-2a67-49eb-af02-7d5285545c77","arxiv_id":"2506.20528","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":7.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":0,"one_line_summary":"Three-dimensional Einstein-Maxwell-scalar theory in a box has five families of Euclidean saddles, including hairy bag-of-gold and boson star-Python's Lunch configurations.","lead":"This paper reports three new families of saddle-point geometries in the thermodynamics of a charged scalar box in three dimensions, adding to the previously known empty and boson star saddles. The result suggests the phase structure of this simple model is as rich as higher-dimensional holographic superconductors, despite the absence of black holes.","discovery_kind":"new_application","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The boson star-PL saddle's patching across r=rmax is demonstrated only at leading order, with no junction-condition or residual checks and an acknowledged zigzag artifact; the existence of this new saddle family is therefore not yet established.","rationale":"The paper's primary new result is the identification of three additional saddle families, of which the boson star-PL is the most exotic and the least supported. The hairless BG saddle is analytic, and the hairy BG saddle is a more standard shooting problem from a horizon. The PL saddle, however, requires a coordinate patching that is asserted but not verified beyond leading order. The absence of code, data, or independent confirmation, combined with the author's own admission of a zigzag artifact and unresolved low-temperature behavior, means the existence claim is currently a numerical suggestion rather than a demonstrated result. The reader's conditional verdict correctly captures this. Our proposed test provides a direct, minimal check that would either confirm the smoothness of the patch or expose a thin-shell defect, and thereby settle whether the PL family is a genuine saddle. If the test passes, the central claim of at least five saddle families would be substantially strengthened; if it fails, the claim reduces to four families, changing the phase diagram.","tokens_in":17798,"tokens_out":11474,"duration_ms":126520,"concrete_test":"Recompute the boson star-PL solution on a single global coordinate chart by writing the metric and matter fields as functions of rho, with rho = -sqrt(rmax-r) on the lower chart and rho = +sqrt(rmax-rtilde) on the upper chart, and integrating the EMS equations (2.5)-(2.8) in rho from the regular center to the boundary at rho = sqrt(rmax-rb). Then compare the field values and first derivatives at rho=0 from this global integration with those obtained from the two-sided shooting described in §4.1.2. If the derivatives do not match to the solver tolerance, or the residual of the equations of motion does not vanish with increased resolution near rho=0, the patched configuration is not a smooth solution and the boson star-PL saddle does not exist.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The central existence claim for the boson star-PL saddle depends on the assertion that the geometry can be smoothly extended through r=rmax by patching two coordinate charts using expansions (4.6)-(4.13). The paper supplies only the leading terms of these expansions and a schematic figure; it does not demonstrate that the next-order terms satisfy the field equations, nor that the induced metric and extrinsic curvature match across the junction. Without such a check, a delta-function curvature (thin shell) at the seam would make the configuration a solution with a brane source rather than a saddle of the EMS action. The paper itself notes a 'small zigzag shape' in the boson star-PL branch near the junction and attributes it to numerical inaccuracy without a convergence study. Because the full solution curve and the free-energy comparison in §4.2 inherit this patching, the existence of the fifth saddle family, and hence the 'many phases' claim, is not yet supported.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper studies Euclidean Einstein-Maxwell-scalar (EMS) theory in three dimensions with a Dirichlet box boundary and no cosmological constant. It claims that, in addition to the known empty and boson-star saddles, the system admits three further Euclidean saddle families: the bag-of-gold (BG) saddle, a hairy generalization of it, and a configuration dubbed the 'boson star-PL' saddle. For q=0 the paper reviews and corrects the thermodynamics of the EM system of Ref. [31]. For finite coupling q it constructs the new saddles by numerical shooting, computes their free energies, and presents phase diagrams in the (T,mu) plane. The central thesis is that the phase structure is as rich as the higher-dimensional AdS cases, with transitions among empty, BG, hairy BG, boson-star, and boson-star-PL phases.","tokens_in":17975,"tokens_out":4431,"duration_ms":48142,"significance":"If the numerical solutions are genuine saddle points, the paper corrects the earlier conclusion of Ref. [33] that the 3D EMS box has only empty and boson-star saddles, and it establishes a concrete low-dimensional realization of bag-of-gold and Python's-lunch-like geometries in gravitational thermodynamics. The explicit expansions (4.1)-(4.17) and the observation of a rescaling symmetry in three dimensions (footnote 12) are useful contributions. The author is also commendably candid about numerical difficulties, including the unresolved low-temperature hairy-BG branch and the acknowledged numerical zigzag in the boson-star-PL branch. However, the central existence claims rest on numerical shooting without code, data, residual checks, or convergence tests, and the boson-star-PL patching is demonstrated only to leading order; these are load-bearing gaps for the paper's main message.","major_comments":[{"comment":"The boson-star-PL saddle is constructed by patching two coordinate charts at r=rmax, but only the leading terms of the expansions are given. The paper does not demonstrate that the full metric and matter fields solve the equations of motion across the seam, nor that the induced metric and extrinsic curvature match there. Without such a junction-condition check, a delta-function (thin-shell) source at r=rmax cannot be excluded, which would make the configuration a solution with a brane source rather than a saddle of the original action. Because the free-energy comparison in Sec. 4.2 and the phase diagram in Figure 10 include this family, the existence of the fifth saddle family is not yet established.","section":"4.1.2, Eqs. (4.6)-(4.13)"},{"comment":"The numerical construction of the boson-star and boson-star-PL families is presented without convergence tests, residual checks, error bars, or code/data availability. The caption of Figure 5 reports 'a small zigzag shape in the boson star-PL branch near the junction' and attributes it to 'a lack of numerical accuracy', while Sec. 4.1.1 notes that 'numerically determining the exact location of the endpoint is a somewhat difficult task'. Since the existence of the new saddle families is the central claim of the paper, these numerical solutions need to be supported by explicit accuracy measures, such as mesh-refinement convergence of the shooting parameters and residual norms of the ODE system.","section":"4.1.1-4.1.2, Figs. 3-5"},{"comment":"The low-temperature behavior of hairy BG saddles is left unresolved: the author states 'I could not obtain the hairy BG solutions near zero temperature' and the corresponding branches are drawn as dotted curves. Consequently, the phase boundary between the boson-star-PL phase and the hairy BG phase is undetermined in the low-T/high-mu region, and the speculation about a possible 'zero temperature hairy BG' is unsupported. The claim that the system has 'many phases' with a phase diagram similar to higher-dimensional cases is therefore incomplete in the region where the new hairy phase is expected to dominate.","section":"4.2, Figs. 9-10"},{"comment":"The thermodynamic stability analysis relies on the assumption that one-loop corrections make the heat capacity of the BG saddle nonnegative, as stated: 'I assume this to be true'. At zero loop, the heat capacity of the BG saddle is exactly zero, so the identification of stable phases and the Hawking-Page-type transition temperature in Sec. 3.2 and Fig. 2 depend on this unproven assumption. This should be flagged explicitly as an assumption in the conclusions so that the phase diagrams of Fig. 10 are not overinterpreted as established zero-loop results.","section":"3.2"}],"minor_comments":[{"comment":"The caption lists the parameter values as '√Gq = 1.2 for (right), √Gq = 0.6 for (middle), and √Gq = 0.3 for (right)'; this should presumably read '1.2 for (left)' and '0.3 for (right)'.","section":"Figure 6 caption"},{"comment":"Both panels of Figure 10 are labelled '(left)' in the caption; the second should be '(right)'.","section":"Figure 10 caption"},{"comment":"The expression 'q /greaterorsimilar2.625 √G/rb' contains a LaTeX artifact; it should read 'q \\gtrsim 2.625√G/rb'.","section":"4.1.2"},{"comment":"The classification 'nr > 0' and 'nr < 0' for the two classes of saddles uses the symbol n_r without a definition.","section":"2"},{"comment":"The numerical evaluation of the free energies for the new saddle families is not described; the paper should specify how the on-shell action is computed numerically and how boundary terms or subtractions are handled.","section":"4.2"}],"recommendation":"major_revision","confidential_remarks":"The paper is within scope and the central idea is interesting, but the existence of the new saddle families hinges on numerical work that is not currently reproducible. I would recommend asking the author to supply code/data and convergence checks, and to either prove smoothness of the PL patching or provide a clear error estimate. The unresolved low-temperature hairy-BG branch should also be addressed by either resolving it or explicitly marking it as an open limitation."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Two things you should know. First, the paper does find something genuinely new: for the 3D EMS system in a Dirichlet box, the known boson-star and empty saddles are not the whole story. The hairy BG saddle and the so-called boson star-PL saddle are not in [33], and the author also corrects an error in his own earlier [31] concerning the BG free energy and energy. That correction is honest and matters for the q=0 phase diagram. Second, the new \"many phases\" claim is still conditional. The boson star-PL saddle is constructed by patching two coordinate charts at r=rmax using leading-order expansions (4.6)-(4.13). No junction-condition check, no residual test, and no independent numerical confirmation are provided; the author himself notes a zigzag in the free-energy curve and attributes it to numerical inaccuracy. Without the next-order terms or a convergence study, a thin-shell delta-function curvature at the seam is not excluded. That is the load-bearing soft spot.\n\nWhat the paper does well: the thermodynamic setup is careful, the expansions near the center and horizon are given, and the numerical shooting curves are plausible. The author is explicit about the low-temperature limits he could not reach and about the disagreement with [33] over the q threshold for boson-star stability. The self-citations are used as background, not fitted to data, and there is no circularity.\n\nThe soft spots, in proportion: the missing numerics are the main issue. This is not a wrong formula; it is a matter of evidence. The hairy BG existence seems better supported by the horizon expansion plus shooting, though its low-temperature behavior is unresolved and the stability language assumes one-loop corrections. The boson star-PL branch is weaker because the patching is only leading-order. These are fixable with code, data, and a junction/residual check.\n\nIf I were the editor, I would send this to a referee rather than desk-reject. The claim is important enough within the subfield, and the author has a track record. The referee should ask for the numerics and a demonstration that the patching is smooth. The paper should probably be published, if at all, as a modified version that answers those points.\n\nFor whom: people working on gravitational thermodynamics with Dirichlet boundaries, Euclidean saddles, and bag-of-gold/Python's-lunch constructions. The broad AdS/CFT audience will wait until the numerics are public.","headline":"Genuinely new saddle families and an honest correction of an earlier error, but the boson star-PL branch needs numerical back-up before the 'many phases' claim is solid.","tokens_in":18483,"tokens_out":2211,"would_cite":false,"duration_ms":24413,"reading_group":"maybe","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":[],"pacs":[],"model":"deepseek-v4-flash","headline":"This paper claims that the three-dimensional Einstein-Maxwell-scalar system in a finite Dirichlet box has five Euclidean saddle families — empty, bag-of-gold, hairy bag-of-gold, boson star, and boson star-Python's-lunch — rather than the…","keywords":["Einstein-Maxwell-scalar system","three-dimensional gravity","bag of gold saddle","boson star","Dirichlet boundary conditions","Euclidean quantum gravity","gravitational thermodynamics","phase transitions"],"falsifier":"Integrate the full field equations (2.5)–(2.8) for a claimed boson star-PL solution using an independent high-precision method, such as spectral collocation, across $r = r_{\\max}$, and verify that the patched fields satisfy the equations to machine precision and that the metric is smooth with all curvature invariants finite; if the residuals do not vanish, or if the zigzag persists as resolution increases, the boson star-PL family is a numerical artifact rather than a true saddle.","tokens_in":17559,"feed_emoji":"🌡️","tokens_out":10282,"duration_ms":96589,"temperature":0.7,"pith_summary":"The paper argues that gravitational thermodynamics in three dimensions with a finite Dirichlet box is as rich as in higher dimensions, despite the absence of black holes. It claims that, on top of the empty (flat-space) and boson star saddles, three further saddle families exist: the bag-of-gold (BG) saddle, its hairy generalization, and a configuration called the boson star-PL (Python's-lunch) saddle. If correct, the phase diagram contains thermal phase transitions between empty, BG, hairy BG, boson star, and boson star-PL phases, with the BG and its hairy version playing the role that black holes play in anti-de Sitter space. The reason to care is that earlier work concluded the phase structure was trivial, whereas this paper says it is not.","feed_headline":"Three new phases found in a 3D gravity box","feed_subtitle":"Bag-of-gold and lunch-shaped saddles join boson stars to reshape the phase diagram.","key_machinery":"The central object is the Euclidean saddle itself, classified by the position of the bolt (the fixed point of the Euclidean time circle) relative to the Dirichlet boundary at $r = r_b$. In ordinary black-hole saddles the bolt lies inside the box; in BG saddles it lies outside; in boson star-PL saddles the metric function $f(r)$ vanishes at a maximum radius $r_{\\max}$ where the coordinate system breaks down, and the geometry is continued beyond $r_{\\max}$ by a second coordinate patch with the expansions (4.6)–(4.13). The argument is carried by shooting from the center or horizon with two free parameters ($a_0, \\phi_0$) or ($a_{G1}, \\phi_{G0}$), imposing the boundary condition $\\phi(r_b) = 0$, and using the one-parameter solution curves so obtained to compute on-shell free energies and identify dominant saddles.","core_discovery":"The central claim is that the Euclidean Einstein-Maxwell-scalar system in three dimensions without a cosmological constant, enclosed in a finite Dirichlet box, admits at least five families of saddle-point solutions, not the two (empty and boson star) identified in prior work. The three additional saddles are: the bag-of-gold (BG) saddle, whose bolt lies outside the box and whose horizon area exceeds the boundary area; the hairy BG saddle, which adds a charged scalar condensate outside the box; and the boson star-PL saddle, a two-coordinate-patch geometry in which the radial coordinate reaches a maximum radius $r_{\\max}$ beyond the boundary, at which point the metric is regular and can be continued back to the boundary. The paper argues that these saddles produce a phase diagram with first- and second-order transitions, including situations where the boson star-PL and hairy BG phases dominate, and that the structure closely parallels higher-dimensional holographic superconductors.","pith_inferences":["A direct numerical search in the four-dimensional Einstein-Maxwell or Einstein-Maxwell-scalar system with a Dirichlet box should also find bulge-type saddles analogous to the boson star-PL branch; the paper only speculates about higher dimensions.","The zero-loop heat capacity of BG saddles vanishes, and the paper assumes one-loop corrections make it positive; this assumption can be tested by computing the one-loop determinant, and a negative result would remove the thermodynamic interpretation offered.","The reported numerical zigzag at the junction of the boson star and boson star-PL curves suggests the two families may be connected by a near-crossing of solution branches rather than a smooth merger; a careful continuation in the shooting parameters could reveal an additional saddle or a change in transition order.","The claimed threshold $q \\simeq 2.625\\sqrt{G}/r_b$ for boson star stability, which corrects an earlier value of about $3.5$, should be verifiable by an independent grid search in $(a_0, \\phi_0)$ with higher shooting precision, since the endpoint region is numerically delicate."],"forward_implications":["The previously claimed simplicity of the three-dimensional EMS phase diagram is wrong; the full diagram includes BG, hairy BG, and boson star-PL phases in the high-chemical-potential, low-temperature region.","The BG and hairy BG saddles provide a finite-entropy, horizon-like high-temperature phase in a system with no black holes, so thermal phase transitions occur without black holes.","The boson star-PL saddle shows that Python's-Lunch-type geometries, with a bulge beyond the boundary, can be the dominant Euclidean saddles, so such geometries must be included in Dirichlet-box gravitational path integral computations.","The phase diagrams match qualitatively those of holographic superconductors, suggesting that the three-dimensional Dirichlet-box EMS system is a toy model for holographic superconductivity without an AdS boundary.","For sufficiently small scalar coupling $q$, the ordinary boson star is never thermodynamically stable and the boson star-PL phase replaces it, so the stability of solitonic phases depends on $q$ in a way not previously recognized."],"supporting_citations":[{"why":"prior work that classified only empty and boson star saddles, the claim this paper corrects","marker":"[33]"},{"why":"introduced the bag-of-gold saddle in the Einstein-Maxwell system and its thermodynamics, which the paper extends to the hairy case","marker":"[31]"},{"why":"no-black-hole theorem in three-dimensional gravity that forces the high-temperature phase to be a BG rather than a black hole","marker":"[38]"},{"why":"Euclidean action with Dirichlet boundary conditions that defines the gravitational path integral ensemble used here","marker":"[10]"},{"why":"grand canonical ensemble for a charged system with Dirichlet boundary, supplying the framework for thermodynamic quantities and free energy","marker":"[28]"},{"why":"holographic superconductor construction that provides the comparison point for the phase structure of the hairy system","marker":"[12]"}],"fun_headline_variants":["Hairy box gravity gains three new phases","3D gravity box now hosts five families of saddles","Bag-of-gold and boson-star-PL enrich 3D box phases","Five saddle families found in a hairy 3D box","Hairy box phases multiply from two to five"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The load-bearing premise is that the numerical shooting solutions, especially the boson star-PL saddle, are genuine smooth saddle points of the Euclidean action: the construction patches two coordinate charts at $r = r_{\\max}$ using low-order expansions and a figure, with no convergence tests, residual checks, or independent confirmation, and the paper itself notes a numerical zigzag at the junction.","fun_headline_variants_meta":{"raw":{"variants":["Hairy box gravity gains three new phases","3D gravity box now hosts five families of saddles","Bag-of-gold and boson-star-PL enrich 3D box phases","Five saddle families found in a hairy 3D box","Hairy box phases multiply from two to five"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000639,"raw_usage":{"total_tokens":2916,"prompt_tokens":892,"completion_tokens":2024,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":508,"completion_tokens_details":{"reasoning_tokens":1943}},"tokens_in":508,"tokens_out":2024,"duration_ms":15834,"temperature":1.0,"reasoning_tokens":1943,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-06T22:46:18.139764+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Integrate the full field equations (2.5)–(2.8) for a claimed boson star-PL solution using an independent high-precision method, such as spectral collocation, across $r = r_{\\max}$, and verify that the patched fields satisfy the equations to machine precision and that the metric is smooth with all curvature invariants finite; if the residuals do not vanish, or if the zigzag persists as resolution increases, the boson star-PL family is a numerical artifact rather than a true saddle.","supporting_citations":[{"cited_title":"A ha iry box in three dimensions,","cited_arxiv_id":null,"evidence_quote":"prior work that classified only empty and boson star saddles, the claim this paper corrects"},{"cited_title":"Thermodynamics of the 3-dimensional Ei nstein-Maxwell system,","cited_arxiv_id":null,"evidence_quote":"introduced the bag-of-gold saddle in the Einstein-Maxwell system and its thermodynamics, which the paper extends to the hairy case"},{"cited_title":"Charged black hole in a grand canonical ensemble,","cited_arxiv_id":null,"evidence_quote":"grand canonical ensemble for a charged system with Dirichlet boundary, supplying the framework for thermodynamic quantities and free energy"}],"review_version":1}