{"id":"9b4095c0-d738-436c-b5d8-e0b92b279e4d","arxiv_id":"2607.19468","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":7.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":0,"one_line_summary":"Massless QED2 on AdS2 is screened, not confined, when the static potential is defined after subtracting position-dependent probe self-energies.","lead":"A two-dimensional model of electromagnetism on a negatively curved spacetime is shown to screen charges rather than confine them, once position-dependent self-energies of the charges are properly subtracted. The result settles a long-standing contradiction in curved-space quantum field theory and comes with a new lattice method plus tensor-network validation.","discovery_kind":"first_principles","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Eq. (3.22) does not keep V_qbarq bounded as d→∞ along all paths: r1 = A e^{s d/L} with s>ν gives V→−∞, so the screening claim is limit-dependent.","rationale":"The reader's weakest assumption (fixed background, flat-space bosonization) is reasonable but not the most load-bearing: those are assumptions that, if wrong, would change the model rather than invalidate the internal logic. The path-dependence of Eq. (3.22) is a direct internal inconsistency with the abstract's universal claim. The paper defines screening as the binding energy staying finite as the geodesic separation diverges; Eq. (3.22) violates this for a continuum of limits where both charges move to the AdS boundary. This is not a matter of consensus or an external assumption; it is a mathematical property of the paper's own central result. The one-form symmetry argument and DMRG simulations provide supporting evidence that the model may indeed be screened, but they do not test the problematic limit, and the analytical proof as written does not establish the claim. The authors could fix this by explicitly defining the d→∞ procedure (e.g., holding one charge at a reference position) or by showing that the diverging paths correspond to an unphysical normalization. Because the issue is serious but potentially fixable, the reader's CONDITIONAL verdict remains appropriate: the paper should not be fully accepted until the path-dependence is resolved. I therefore keep the verdict unchanged, while disagreeing with the reader's identification of the weakest point.","tokens_in":27898,"tokens_out":15126,"duration_ms":150302,"concrete_test":"Re-evaluate the central claim along the family of trajectories r1 = A e^{s d/L}, r2 = r1 e^{d/L} for s = 0, ν/2, ν, 2ν. Plot V from Eq. (3.22) versus d for each; for s>ν the curve diverges, falsifying 'remains finite as d→∞.' Also repeat in global coordinates using Eq. (3.48) with ρ1 = α d/L, ρ2 = (α+1)d/L, α>ν. If the divergence is confirmed, the paper must either restrict the definition of the d→∞ limit (e.g., fix one charge relative to the boundary) or show that the divergent path is unphysical (e.g., requires infinite proper acceleration). This check settles whether the claimed screening is well-defined under the paper's own diagnostic.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The central claim is that after subtracting position-dependent self-energies, the static q-qbar potential remains finite as the geodesic separation d→∞, establishing screening. Eq. (3.22) gives V_qbarq = −(e^2 q^2/(2ν+1)) r1 (r1/r2)^ν. Since d = L log(r2/r1), this is V = −C r1 e^{-ν d/L}, C = e^2 q^2/(2ν+1). The paper only examines paths with r1 fixed or r2 fixed (Eq. 3.24), but d→∞ permits paths in which both charges move toward the AdS boundary: take r1 = A e^{s d/L}, r2 = A e^{(s+1)d/L}. Then r2/r1 = e^{d/L}, so d is indeed the geodesic separation, and V = −C A e^{(s−ν)d/L}. For any s>ν this diverges to −∞ as d→∞. The same behavior follows in the global frame from Eq. (3.48): with ρ1 = α d/L, ρ2 = (α+1)d/L, the asymptotics a(ρ) ~ e^ρ/2, u'_L ~ e^{νρ}, u'_R ~ e^{-(ν+1)ρ} give V ∝ −e^{(α−ν)d/L}. Thus the subtracted potential is not a function of d alone and does not remain finite under all limiting procedures. The paper's resolution of the earlier confining/screening ambiguity is therefore incomplete: it fixes the two limits considered in Ref. [32], but introduces a third, equally natural limit with divergent binding energy. The one-form symmetry argument may still imply screening, but Eq. (3.22) does not establish it under the paper's own definition. The reader's identified weakest assumptions (fixed background, flat-space bosonization) are external; this concern is internal to the paper's central calculation.","agreement_with_reader":"disagree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper studies massless single-flavor QED_2 on AdS_2, in both Schwarzschild and global frames, and claims to resolve a longstanding ambiguity about confinement versus screening. The main analytic tool is exact bosonization of the massless theory in a fixed AdS_2 background. After computing the ground-state energy in the presence of two external probe charges, the authors subtract position-dependent single-probe self-energies and obtain the qar q potential in closed form, e.g. Eq. (3.22). They conclude that the subtracted potential remains finite as the geodesic separation diverges, so the theory is screened, in agreement with the explicit breaking of the electric one-form symmetry. The same conclusion is extended to the global frame and to finite temperature. The paper also proposes a covariant lattice discretization of curved-space fermions and uses DMRG/tensor-network simulations to support the analytic prediction.","tokens_in":28351,"tokens_out":22422,"duration_ms":198397,"significance":"If the central screening claim is correct under a well-defined prescription, the paper would resolve an apparent contradiction in earlier treatments [31-33] and provide an exactly solvable example of screening in curved spacetime. The covariant discretization of the spin and gauge connections is a useful technical contribution, and the tensor-network validation is a strength: the simulations follow the continuum profile without free normalization. However, the paper's advertised conclusion that the theory screens 'independently of how d is taken to infinity' is not established by the presented calculation, because the subtracted potential (3.22) is path-dependent and can diverge along other equally natural limiting procedures.","major_comments":[{"comment":"The screening claim is not a function of the geodesic separation alone. Equation (3.22) gives V = -C r_1 (r_1/r_2)^ν with d = L log(r_2/r_1). The paper only examines the limits with r_1 fixed or r_2 fixed. But take r_1 = A e^{s d/L}, r_2 = A e^{(s+1)d/L} with s > ν. Then d = L log(r_2/r_1) is exactly the stated geodesic separation, and V = -C A e^{(s-ν)d/L}, which diverges to -∞ as d → ∞. Thus Eq. (3.22) does not remain finite for all ways of sending d to infinity. This directly contradicts the abstract's claim that the potential 'remains finite as the geodesic separation is taken to infinity' and the conclusion in Sec. 5 that screening holds 'independently of how d→∞ is taken.' At minimum, the paper must either prove boundedness along all sequences with d→∞, or explicitly adopt and justify a restricted definition of 'static potential' in curved space. As written, the resolution of the e","section":"Sec. 3.2, Eqs. (3.22)-(3.24)"},{"comment":"The same path-dependence problem appears in the global frame. Using the asymptotics of u_L, u_R at the two boundaries, the global subtracted potential V(ρ_1,ρ_2) = - (π q^2/W) a(ρ_1)a(ρ_2) u'_L(ρ_1)u'_R(ρ_2) contains factors that grow exponentially with ρ_1 + ρ_2. For a sequence with ρ_1 = α d/L and ρ_2 = (α+1)d/L, the potential behaves like -C exp[(2α - ν)d/L] (up to the explicit asymptotic factors in (3.52)), which diverges for sufficiently large α. Hence the claimed frame-independent, procedure-independent screening conclusion is not supported by the analytic result. A revised paper should either restrict to a physically motivated convention for separating the charges and state it as part of the definition of screening, or accept that the static potential in curved space is not a function of d alone and the notion of 'screening' needs refinement.","section":"Sec. 3.3, Eq. (3.48)"}],"minor_comments":[{"comment":"Typo: 'continuuum' should be 'continuum'.","section":"Sec. 1.2"},{"comment":"The heading 'F ree energy' has an unintended space; also 'q¯qpotential' appears with inconsistent spacing throughout.","section":"Sec. 3.4"},{"comment":"For the single-charge self-energy (q,0), the boundary condition E(∞)=0 is compatible with nonzero total charge only because the dynamical fermions carry a compensating vacuum polarization. This is physically clear but deserves a sentence, since the compact-support formula in Eq. (3.18) is used outside its strict domain in deriving Eq. (3.21).","section":"Sec. 3.2, Eq. (3.17)"},{"comment":"The numerical section would be strengthened by reporting truncation errors, bond-dimension convergence, and finite-size extrapolation for the DMRG data. The qualitative agreement is convincing, but quantitative claims would benefit from controlled error estimates.","section":"Sec. 4"}],"recommendation":"major_revision","confidential_remarks":"The core technical computations appear internally consistent, and the lattice discretization is a potentially valuable contribution. The main obstacle is the central screening claim: Eq. (3.22) does not establish boundedness for all limiting procedures, despite the abstract and conclusions claiming otherwise. The authors may be able to fix this by carefully defining what 'geodesic separation going to infinity' means for a non-translation-invariant background, and by acknowledging that the static potential depends on the pair's absolute positions. If that definitional issue is resolved, the paper could be acceptable; as it stands, the advertised resolution of the earlier ambiguity is too strong."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Short version: this paper has a genuine insight and a useful technical contribution, but the advertised resolution of the screening-versus-confinement ambiguity does not survive contact with its own Eq. (3.22).\n\nWhat is good: the observation that probe self-energies in curved space are position-dependent and must be subtracted before reading off a binding potential is correct and explains why earlier work [31–33] found contradictory, limit-dependent behavior. The bosonized calculation is exact for the massless theory because the action is quadratic; the Green's function, on-shell energy, and flat-space limit all check out. The covariant discretization, packaging spin and gauge connections into one Hermitian difference operator, is a genuine fix to the separate non-Hermiticity problem in [37], and the DMRG electric field profile matches the continuum curve with no free normalization. Those parts deserve credit.\n\nWhere it falls down: the paper claims the subtracted potential remains finite as d→∞ 'independently of how the geodesic separation is taken to infinity.' That is not what Eq. (3.22) says. With d = L log(r2/r1), choose r1 = A e^{s d/L}, r2 = A e^{(s+1)d/L}. This is a perfectly good path with geodesic separation d, and then V = -C A e^{(s-ν)d/L}, which diverges to -∞ for any s > ν. The global-frame expression (3.48) shows the same behavior. So the screening conclusion is established only for the two specific limits the authors plot (r1 fixed or r2 fixed), not for all limits. Since the paper's entire resolution of the prior ambiguity is about eliminating limit-dependence, this is a load-bearing soft spot, not a cosmetic one.\n\nTwo more caveats, in proportion. The flat-space bosonization dictionary is assumed on AdS2 without a derivation; curvature corrections would change the Green's function and the potential. And the background is treated as fixed, which the authors acknowledge. Neither is fatal, but both limit how strongly the result can be stated. No code or data is released, so the numerics are hard to reproduce independently.\n\nWho this is for: people working on the Schwinger model in curved space, lattice gauge theory with fermions in curved backgrounds, and tensor-network simulations of gauge theories. It deserves a serious referee: the technical core is substantial and the discretization alone is worth engaging with. But the referee should push hard on the path-dependence of the screening claim. As written, I would not accept the central conclusion without a clear physical definition of which limiting procedure counts, or a proof that the self-energy-subtracted potential depends only on d after a well-motivated regulator.","headline":"The self-energy subtraction is a real advance, but the paper's central claim that the screened potential is independent of how the geodesic separation is taken to infinity is false on its own formula.","tokens_in":28809,"tokens_out":2709,"would_cite":false,"duration_ms":33912,"reading_group":"yes","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":[],"pacs":[],"model":"deepseek-v4-flash","headline":"Massless quantum electrodynamics on anti-de Sitter space screens rather than confines, once the position-dependent self-energies of the probe charges are subtracted from the static potential.","keywords":["Schwinger model","QED2","AdS2","confinement","screening","bosonization","static potential","tensor networks"],"falsifier":"Compute the same subtracted binding potential while allowing the probe energies to backreact on the geometry (for instance by solving the dilaton-gravity constraints); if it grows with separation rather than following the predicted exponential decay e^{-νd/L}, the screening claim fails. Alternatively, run a lattice simulation at much smaller fermion mass and much larger separation, taking the continuum limit, and check that the subtracted potential continues to decay rather than plateauing at a nonzero value.","tokens_in":27807,"feed_emoji":"⚛️","tokens_out":9021,"duration_ms":85526,"temperature":0.7,"pith_summary":"The paper takes up a long-standing puzzle: the massless Schwinger model on AdS2 had been reported as both confining and screening, depending on how the charge-anticharge separation was sent to infinity. It argues that the earlier readings were artifacts of identifying the full ground-state energy with the binding potential, because in curved space the one-body self-energy of each probe is position dependent. After subtracting those self-energies, the remaining binding potential is obtained in closed form and decays exponentially with geodesic separation, in both the Schwarzschild and global frames and at finite temperature. Tensor-network simulations of the same lattice Hamiltonian reproduce the screened potential, the electric-field profile, and real-time string breaking.","feed_headline":"Screens, not confines: massless curved-space QED","feed_subtitle":"On AdS2, subtracting position-dependent single-probe self-energies reveals a binding potential that decays to zero with separation.","key_machinery":"The analytic argument runs on bosonization: the massless Schwinger model is mapped to a massive dual scalar, and the static potential is carried by the Green's function of the operator -∂_r(f ∂_r φ) + (e²/π)φ with f = r²/L², whose homogeneous solutions are the powers r^ν and r^{-ν-1}. The conceptual move that resolves the puzzle is subtracting the position-dependent one-body self-energies before reading off the binding energy. For the numerical half, the paper packages the kinetic term together with the spin and gauge connections into a single Hermitian generalized covariant derivative, i∇ = i[a(r)(∂_r + iA_r) + a'(r)/2], before discretizing; this guarantees Hermiticity at finite lattice spa","core_discovery":"The paper's central claim is that single-flavor massless QED2 on AdS2, with or without an AdS2 black hole, is screened rather than confined. The static potential between an opposite-charge probe pair, after subtracting each probe's position-dependent self-energy, is V = -(e²q²/(2ν+1)) r1 (r1/r2)^ν in Schwarzschild coordinates, with ν = -1/2 + sqrt(1/4 + e²L²/π). As the geodesic separation d = L log(r2/r1) goes to infinity this tends to zero exponentially, no matter which charge is sent to infinity. The same conclusion is reached in global AdS2 coordinates and, at finite temperature, from the free energy after subtracting one-body shifts. This agrees with the explicit breaking of the U(1) ele","pith_inferences":["The paper notes that it treats the AdS2 geometry as fixed; if gravitational backreaction of the probes and fields were included, the horizon and effective potential would shift, and the screening verdict could in principle change.","The same self-energy-subtraction prescription could be applied to other curved backgrounds, such as dS2 or near-horizon geometries, where past work reported region-dependent confinement; if the bosonized scalar's Green's function has normalizable modes, a similar screened phase would be expected.","A curvature-expanded check of the bosonization dictionary, testing corrections to the flat-space current-to-scalar map on AdS2, would show whether the exponential decay formula is exact or only leading order.","The covariant discretization recipe may transfer to lattice gauge theories in higher dimensions on curved slices, where separate discretization of kinetic and spin-connection terms breaks Hermiticity."],"forward_implications":["The earlier frame- and limit-dependent claims of confinement in curved-space QED2 are resolved: after self-energy subtraction, the binding potential saturates regardless of how the geodesic separation is taken to infinity.","No linear string tension appears between the probes; the electric field inside the pair decays exponentially with the coupling, and exterior screening-cloud tails decay with geodesic distance.","Finite temperature does not restore confinement: the subtracted q-qbar potential, built from a Hartle-Hawking Green's function, remains finite and decays with separation.","The flat-space Schwinger model is recovered as the AdS2 radius goes to infinity in both frames, giving the familiar Yukawa-screened potential as a limit of the curved-space result.","Real-time tensor-network evolution shows an initially confined e+e- pair connected by a Wilson line breaking through pair production, confirming the screened phase nonperturbatively."],"fun_headline_variants":["Massless QED2 on AdS2: screened, not confined","AdS2 black hole doesn't confine massless charges","Self-energy subtraction shows AdS2 QED screens","Curved-space QED2: probes shielded at infinity","Screening wins: massless QED2 on AdS2"],"cache_read_input_tokens":2304,"weakest_assumption_plain":"The analysis assumes the AdS2 geometry is fixed and does not react to the probe charges or quantum fields, and that the standard flat-space rule expressing fermion currents as derivatives of a scalar field remains valid on curved AdS2; if either assumption fails, the Green's function and the resulting screening verdict change.","fun_headline_variants_meta":{"raw":{"variants":["Massless QED2 on AdS2: screened, not confined","AdS2 black hole doesn't confine massless charges","Self-energy subtraction shows AdS2 QED screens","Curved-space QED2: probes shielded at infinity","Screening wins: massless QED2 on AdS2"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000312,"raw_usage":{"total_tokens":1694,"prompt_tokens":908,"completion_tokens":786,"prompt_tokens_details":{"cached_tokens":256},"prompt_cache_hit_tokens":256,"prompt_cache_miss_tokens":652,"completion_tokens_details":{"reasoning_tokens":701}},"tokens_in":652,"tokens_out":786,"duration_ms":8715,"temperature":1.0,"reasoning_tokens":701,"cache_read_input_tokens":256,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-01T12:38:38.692393+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Compute the same subtracted binding potential while allowing the probe energies to backreact on the geometry (for instance by solving the dilaton-gravity constraints); if it grows with separation rather than following the predicted exponential decay e^{-νd/L}, the screening claim fails. Alternatively, run a lattice simulation at much smaller fermion mass and much larger separation, taking the continuum limit, and check that the subtracted potential continues to decay rather than plateauing at a nonzero value.","supporting_citations":[],"review_version":1}