{"id":"66033fbd-335c-49db-b763-207495be4923","arxiv_id":"hep-th/0303116","paper_version":1,"verdict":"ACCEPT","confidence":"HIGH","novelty_score":8.0,"correctness_risk":"low","formal_verification":"none","parameter_count":0,"one_line_summary":"The DGP brane-world model loses predictivity below a length scale of roughly 1000 km, set by a brane-bending scalar mode whose kinetic term arises only through mixing with the graviton.","lead":"The DGP model makes gravity four-dimensional on a brane and five-dimensional far away. This paper shows that the same model becomes strongly interacting, and therefore unpredictable, at a distance of order 1000 km when the crossover scale is the Hubble length.","discovery_kind":"first_principles","skeptic_critique":{"model":"selfhosted/deepseek-v4.1-flash","headline":"No significant objection identified: the λ3 scale follows from the explicit cubic vertex in Eq. 3.38 with a fixed coefficient, so the reader's tuning concern does not land.","rationale":"The reader's verdict of ACCEPT with high confidence is correct, but the reader's weakest_assumption is misdirected. The reader worries that tuned boundary counterterms could raise the cutoff above λ3, by analogy with the massive-gravity tuning that lifts Λ5 to Λ3. That analogy does not transfer: in massive gravity the tuning raises the scale from Λ5 to Λ3, but Λ3 is the floor because the cubic Goldstone term is present; in DGP the scale is already Λ3, and the cubic term is a fixed tree-level prediction of the bulk action, not a free counterterm. The paper's gauge-choice argument (paragraph after Eq. 3.31) and non-renormalization argument (Eq. 3.42) close the only routes by which the cubic term could be removed. The explicit computation of Eq. 3.38, together with the canonical normalization from Eq. 3.24, gives Λ = (M4 m²)^{1/3} directly, independent of any assumption about higher-dimension operators. The secondary results are also sound: the negative-energy solution of Section 4 is a clean Birkhoff and junction computation, and the observation that it sits at curvature of order m², the same scale where the derivative expansion fails, is a genuine consistency argument. The curved-background section is the weakest part, as the reader notes, because Eq. 5.22 rests on the small-wavelength replacement Eq. 5.13, which is justified by a flat-space limit and checked explicitly only for de Sitter; but this is a secondary result and does not bear on the central λ3 claim. The paper is candid about what it does not settle: Section 6 states that any statement below R* requires assumptions about the UV completion of the π sector. The one genuine blemish is the printed power-counting formula Eq. 3.30, which for p=0, q=3 gives (m²M4)^{1/7} rather than the (m²M4)^{1/3} of Eq. 3.31, indicating a typo in the exponent. This typo does not affect the central claim, because the explicit cubic computation in Eq. 3.38 and the hierarchy M5 ≫ Λ independently establish that the cubic term is the leading strong interaction. I would not change the verdict; the paper's central claim holds up under scrutiny, and the reader's tuning concern does not land.","tokens_in":17603,"tokens_out":14925,"duration_ms":274546,"concrete_test":"Compute the on-shell three-point amplitude from the cubic vertex in Eq. 3.38 after canonical normalization, and verify that it scales as 1/Λ³ with Λ = (M4 m²)^{1/3}. If the amplitude is nonzero and scales this way, the strong-coupling scale is λ3 and cannot be raised by tuned boundary counterterms; if it vanishes or scales differently, the reader's tuning concern would land. As a secondary check, recompute the exponent in Eq. 3.30 for p=0, q=3 and confirm whether it reproduces Eq. 3.31; a mismatch would confirm the typo but would not change the central claim.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The central claim is that DGP becomes strongly coupled at λ3 ~ (m²M4)^{1/3}. The derivation does not rest on an assumption that all boundary operators enter with O(1) coefficients. The cubic vertex ΔL = M4² m (∂π)²□π in Eq. 3.38 is a tree-level prediction of the bulk Einstein action, not a tunable counterterm. Once the π kinetic term 3M4²m²π□π (Eq. 3.24) is fixed, the canonical normalization π̂ = √6 M4 m π is fixed, and the cubic coupling is 1/(M4 m²) = 1/Λ³ with Λ = (M4 m²)^{1/3}. No choice of higher-dimension boundary counterterms can remove this term, because it is present at tree level with a coefficient determined by the bulk. The paper explicitly addresses the only escape route the reader raises: a nonlinear change of the de Donder gauge condition cannot cancel the cubic term (paragraph following Eq. 3.31), and the non-renormalization argument at Eq. 3.42 shows that loops do not renormalize it. The massive-gravity analogy the reader invokes actually supports the paper: in massive gravity the tuning raises the scale from Λ5 to Λ3, but Λ3 is the floor because the cubic term is present; in DGP the scale is already Λ3, so there is no room to raise it further by tuning. The one genuine blemish is that the printed power-counting formula Eq. 3.30, evaluated at p=0, q=3, gives (m²M4)^{1/7} rather than the (m²M4)^{1/3} of Eq. 3.31, indicating a typo in the exponent. This typo does not affect the central claim, because the explicit cubic computation in Eq. 3.38 and the hierarchy M5 ≫ Λ independently establish that the cubic term is the leading strong interaction.","agreement_with_reader":"disagree"},"referee_report":{"model":"selfhosted/deepseek-v4.1-flash","summary":"The paper argues that the DGP model becomes strongly coupled at a length scale λ3 ~ (λ_DGP^2/M4)^{1/3}, which is about 1000 km when λ_DGP is of order the Hubble length. The strongly coupled mode is identified as a longitudinal brane-bending Goldstone field π that solves the linearized bulk equations, has H_μν = 0, and obeys de Donder gauge (Eq. 3.26); its kinetic term arises only from mixing with transverse graviton polarizations (Eq. 3.24). The cubic vertex M4^2 m ∂π∂π□π is computed explicitly in Eq. (3.38), giving Λ3 = (m^2 M4)^{1/3}. The paper also presents a negative-energy classical solution and argues it lies at the edge of the effective theory's validity, and it analyzes curved backgrounds, finding that de Sitter curvature makes π a ghost and that a DGP term on the IR brane of Randall-Sundrum makes the radion a ghost.","tokens_in":17905,"tokens_out":13416,"duration_ms":231677,"significance":"This is an important and influential result. The derivation is explicit and parameter-free: the scale follows from the bulk Einstein action and the DGP kinetic term without fitting to data, and the prediction that gravity loses predictivity below roughly 1000 km is falsifiable. The identification of the strongly coupled mode with a brane-bending Goldstone field, the explicit cubic vertex in Eq. (3.38), and the non-renormalization argument at Eq. (3.42) are all substantive contributions. The extension to curved backgrounds and to Randall-Sundrum is a useful bonus.","major_comments":[],"minor_comments":[{"comment":"Evaluating the displayed power-counting formula at p=0, q=3 gives (m^2 M4)^{1/7}, whereas Eq. (3.31) and the explicit computation in Eq. (3.38) give (m^2 M4)^{1/3}. The discrepancy arises because the schematic derivative counting in Eq. (3.28) assigns one derivative per field, while the leading cubic term ∂π∂π□π carries four derivatives; the formula should be corrected or the counting convention stated explicitly.","section":"Section 3, Eq. (3.30)","authors_acknowledge":null,"authors_acknowledge_p":null},{"comment":"The phrase 'at the same scale scale λ3' duplicates the word 'scale' and should read 'at the same scale λ3'.","section":"Abstract","authors_acknowledge":null,"authors_acknowledge_p":null},{"comment":"The sentence 'Because the scale R* is very large for astrophysical sources, it appeas that the DGP model may describe our universe [2]' contains a typo: 'appeas' should be 'appears'. The same paragraph also has 'This property suggest' where 'suggests' is intended.","section":"Section 1","authors_acknowledge":null,"authors_acknowledge_p":null},{"comment":"The word 'Minkowsky' should be 'Minkowski'.","section":"Section 5.1","authors_acknowledge":null,"authors_acknowledge_p":null},{"comment":"The estimate that the strong-coupling scale on the Planck brane is (M5 k)^{1/2} is used to support the holographic interpretation but is not derived; a short derivation or a reference would make this step checkable.","section":"Section 5.3","authors_acknowledge":null,"authors_acknowledge_p":null},{"comment":"The notation ∂(N)^p(∂π)^q is ambiguous about whether the derivatives act on individual fields or on the product; please define the convention, since it feeds directly into the power-counting formula of Eq. (3.30).","section":"Section 3, Eq. (3.28)","authors_acknowledge":null,"authors_acknowledge_p":null},{"comment":"Reference [8] is cited only as arXiv:hep-th/0210184; if it was unpublished at the time of writing, that should be noted so readers can judge the status of the massive-gravity comparison.","section":"Reference [8]","authors_acknowledge":null,"authors_acknowledge_p":null}],"recommendation":"accept","confidential_remarks":null},"author_rebuttal":null,"desk_editor":{"model":"selfhosted/deepseek-v4.1-flash","letter":"Bottom line: this is the paper that pins down where DGP stops being predictive, and it does it by computing rather than by analogy. The λ3 ~ (m²M4)^{1/3} claim in the abstract is earned in Sections 2 and 3, not asserted, and that is the difference between this and the earlier hand-waving about running the massive-gravity cutoff into |p|/λDGP (which the authors themselves flag in Section 1 as a heuristic). Worth your time if you care about DGP, massive gravity, or modified-gravity model building generally. The 1000 km number is the one everybody quotes, and it holds up here. What is genuinely new is the machinery: the boundary effective action with free boundary conditions (Section 2, Section 3), the identification of the strongly coupled mode as a brane bending that is pure gauge in the bulk, has H_μν = 0, and obeys de Donder (Eq. 3.26), the diagonalization that shows π gets its kinetic term only through mixing (Eq. 3.24), and the explicit cubic vertex M4² m ∂π∂π□π in Eq. 3.38. The non-renormalization check at Eq. 3.42 is a nice touch and it is correct. Section 4 is a clean Birkhoff-plus-junction computation, and the observation that the negative-energy solution sits at curvature of order m², exactly where the derivative expansion dies, is a real consistency argument rather than rhetoric. Soft spots, in order of size. First, Eq. 3.30 does not reproduce Eq. 3.31: put p = 0, q = 3 into the printed exponent and you get (m²M4)^{1/7}, not ^{1/3}. That is a typo in the general power-counting formula, not in the physics, since Eq. 3.38 fixes the cubic coefficient directly. Still, it is the formula people will copy, so it deserves an erratum. Second, Section 5 is the thinnest part. Eq. 5.22 leans on the small-wavelength replacement Eq. 5.13, justified by a flat-space limit and checked explicitly only for de Sitter. The RS IR-brane ghost claim is cross-checked against the radion kinetic term Eq. 5.43, which is reassuring, but the general curved case is sketched. On the tuning worry in your report: I don't think it lands. The cubic is tree level from the bulk Einstein action, canonical normalization is fixed by Eq. 3.24, so the coupling is 1/Λ³ with no freedom to dial. The massive-gravity analogy cuts the other way, since there Λ3 is the floor. Send it out. This is a serious paper and the referees should engage with it.","headline":"The paper that actually computes the DGP strong-coupling scale rather than guessing it, with one printed formula that doesn't reproduce its own result.","tokens_in":18673,"tokens_out":1588,"would_cite":true,"duration_ms":27588,"reading_group":"yes","serious_thinker":"yes","would_accept_peer_review":true},"pith_extraction":{"msc":[],"pacs":[],"model":"selfhosted/deepseek-v4.1-flash","headline":"The DGP braneworld is strongly interacting at a length scale $�lambda_3 �sim (�lambda_{�rm DGP}^2/M_4)^{1/3}$, about 1000 km for a Hubble-scale crossover.","keywords":["DGP model","braneworld gravity","strong coupling","Goldstone mode","massive gravity","vDVZ discontinuity","de Sitter background","negative energy solutions"],"falsifier":"A calculation of the four-point amplitude for the brane-bending mode $�pi$ that shows it stays perturbative well above $�Lambda=(m^2M_4)^{1/3}$ because of a natural boundary counterterm would falsify the claimed cutoff; conversely, observing unitarity violation at that scale would confirm it.","tokens_in":18860,"feed_emoji":"🌌","tokens_out":6382,"duration_ms":118936,"temperature":0.7,"pith_summary":"The paper asks whether the DGP braneworld—a 5D bulk with a 4D Einstein term on a brane—really behaves as a weakly coupled 4D gravity theory at distances shorter than its crossover scale $�lambda_{�rm DGP}$. It argues that it does not: a longitudinal brane-bending mode, the Goldstone field eaten by the graviton, has no independent kinetic term at leading order and gets one only by mixing with transverse graviton polarizations. That small kinetic term makes the mode strongly interacting at $�lambda_3 �sim (�lambda_{�rm DGP}^2/M_4)^{1/3}$, which is about 1000 km if $�lambda_{�rm DGP}$ is of order the Hubble length. The paper also finds negative-energy classical solutions at the edge of the effective theory and shows that positive curvature can turn the Goldstone mode into a ghost. The stake is whether DGP can be a predictive alternative to dark energy, or whether it requires new physics at a short distance that is still far larger than the Planck length.","feed_headline":"DGP gravity loses predictivity below 1,000 km","feed_subtitle":"A brane-bending Goldstone mode becomes strongly coupled, so the model's 4D gravity description fails at short distances.","key_machinery":"The argument rests on a boundary effective action for 5D gravity with no a priori boundary condition. Splitting the metric into lapse, shift, and 4D metric, fixing de Donder gauge in the bulk, and solving the bulk equations with fixed boundary values produces a boundary action whose residual gauge freedom is parameterized by a brane-bending shift $�Xi^5=e^{-y�Delta}�pi$. This mode is pure gauge in the bulk, leaves the induced 4D metric unchanged at linear order, and therefore has no large DGP kinetic term; its kinetic term arises only from mixing with the transverse graviton, which is the technical origin of the strong coupling. The same formalism yields the cubic interaction and, on curved backgrounds, the curvature-dependent $�pi$ kinetic term that controls the ghost. A separate 5D Schwarzschild solution with an $S^3$ boundary supplies the negative-energy example and shows that the instability sits at curvature $O(m^2)$.","core_discovery":"The central claim is that DGP has a hidden strong-coupling scale $�lambda_3 �sim (�lambda_{�rm DGP}^2/M_4)^{1/3}$, not just the crossover scale $�lambda_{�rm DGP}$ at which gravity becomes five-dimensional. The paper derives this by integrating out the bulk with free boundary conditions and organizing the boundary action around the brane-bending mode $�pi$; at quadratic order $�pi$ has kinetic term $3M_4^2 m^2 �pi�,�Box�pi$ (Eq. (3.24)), suppressed by $m=M_5^3/M_4^2$, and at cubic order the explicit interaction $M_4^2 m�,�partial_�mu�pi�,�partial^�mu�pi�,�Box�pi$ (Eq. (3.38)) makes the mode strongly coupled at $�Lambda �sim (m^2M_4)^{1/3}$. For $�lambda_{�rm DGP}$ of order the Hubble length this gives $�lambda_3 �sim 1000$ km, below which the effective theory loses predictivity. The paper adds a negative-energy classical solution satisfying the dominant energy condition, confined to curvature of order $m^2$, and shows that in de Sitter-like backgrounds the $�pi$ kinetic term becomes $3M_4^2 m(m-2C)�pi�,�Box�pi$, so the mode is a ghost when the background curvature $C$ reaches $m/2$.","pith_inferences":["The paper's power counting treats all allowed boundary operators as entering at the cutoff; a natural next step is to ask whether DGP admits a tuned counterterm analogous to the massive-gravity tuning that raises $�Lambda_5$ to $�Lambda_3$, which would weaken the 1000 km claim to a statement about a particular matching.","Because the ghost appears for positive curvature, a consistent UV completion of DGP self-acceleration must either remove the ghost or forbid the background; this links the model's cosmological use to the same curved-space stability questions that arise in massive gravity.","The negative-energy solutions suggest a general relationship between dominant-energy brane matter and negative 5D mass in braneworlds; a scan of the allowed equation-of-state and mass range could show whether the $�lambda_3$ cutoff is necessary or merely sufficient to exclude them."],"forward_implications":["If $�lambda_{�rm DGP}$ is of order the Hubble length, DGP predictions for gravity at distances below about 1000 km require a UV completion; the model cannot be treated as a complete 4D effective theory there.","The Vainshtein recovery of general relativity around a source at $R_*�sim(R_S�lambda_{�rm DGP}^2)^{1/3}$ occurs where the 4D curvature is $O(m^2)$, so the recovery is not established within the quantum effective theory that the paper analyzes.","A cutoff at $�lambda_3$ removes the negative-energy classical solutions, because those solutions have $|R_S|<1/m$ and curvature $O(m^2)$.","In a de Sitter-like background the Goldstone mode becomes a ghost when $C=m/2$, so the self-accelerating DGP background is unstable in this effective description.","A DGP kinetic term on the Planck brane of Randall-Sundrum does not introduce a new strong-coupling scale below the AdS scale, while a DGP term on the IR brane makes the radion a ghost when $m<2k$."],"supporting_citations":[{"why":"Introduces the DGP model and its crossover scale $�lambda_{�rm DGP}$.","marker":"[1]"},{"why":"Supplies the self-accelerating cosmological motivation for taking $�lambda_{�rm DGP}$ near the Hubble scale.","marker":"[2]"},{"why":"Establishes the vDVZ discontinuity that makes the massless-graviton limit subtle.","marker":"[3]"},{"why":"Gives the Vainshtein nonlinear recovery argument that the paper re-examines at $R_*$.","marker":"[4]"},{"why":"Shows the nonlinear DGP solution approaches general relativity below $R_*$, the result the paper argues lies at the edge of the effective theory.","marker":"[5]"},{"why":"Supplies the massive-gravity Stückelberg and strong-coupling analogy, including the $�Lambda_5$ versus $�Lambda_3$ tuning.","marker":"[8]"},{"why":"Treats massive gravity on curved backgrounds, the analogue for the curvature-dependent ghost.","marker":"[9]"},{"why":"Provides the Randall-Sundrum setup used in the DGP-on-RS section.","marker":"[10]"},{"why":"Gives the Higuchi ghost result for a light graviton in de Sitter space, the analogue for the $�pi$ ghost.","marker":"[14]"},{"why":"Supplies the radion kinetic term used to identify the IR-brane ghost.","marker":"[16]"}],"fun_headline_variants":["DGP strong coupling cuts predictivity at 1,000 km","DGP's brane-bending mode goes strong below 1,000 km","Eaten Goldstone makes DGP strongly coupled at 1,000 km","DGP gravity: negative-energy solution removed by the 1,000 km cutoff","DGP Goldstone becomes a ghost when curvature reaches m/2"],"cache_read_input_tokens":0,"weakest_assumption_plain":"The derivation assumes that every boundary operator allowed by 4D general covariance enters with a coefficient of order one at the cutoff, so that no tuned counterterm raises the strong-coupling scale or gives the Goldstone mode a large independent kinetic term.","fun_headline_variants_meta":{"raw":{"variants":["DGP strong coupling cuts predictivity at 1,000 km","DGP's brane-bending mode goes strong below 1,000 km","Eaten Goldstone makes DGP strongly coupled at 1,000 km","DGP gravity: negative-energy solution removed by the 1,000 km cutoff","DGP Goldstone becomes a ghost when curvature reaches m/2"]},"model":"selfhosted/deepseek-v4.1-flash","effort":"low","attempts":1,"cost_usd":0.0,"raw_usage":{"total_tokens":2084,"prompt_tokens":1035,"completion_tokens":1049,"prompt_tokens_details":null,"completion_tokens_details":{"reasoning_tokens":963}},"tokens_in":1035,"tokens_out":1049,"duration_ms":19686,"image_pages":[],"temperature":1.0,"reasoning_budget":50,"reasoning_tokens":963,"transport_provider":"selfhosted","cache_read_input_tokens":0,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-10-29T17:16:19.667125+00:00","model_set":{"reader":"selfhosted/deepseek-v4.1-flash"},"falsifier":"A calculation of the four-point amplitude for the brane-bending mode $�pi$ that shows it stays perturbative well above $�Lambda=(m^2M_4)^{1/3}$ because of a natural boundary counterterm would falsify the claimed cutoff; conversely, observing unitarity violation at that scale would confirm it.","supporting_citations":[{"cited_title":"4D Gravity on a Brane in 5D Minkowski Space","cited_arxiv_id":"hep-th/0005016","evidence_quote":"Introduces the DGP model and its crossover scale $�lambda_{�rm DGP}$."},{"cited_title":"Accelerated Universe from Gravity Leaking to Extra Dimensions","cited_arxiv_id":"astro-ph/0105068","evidence_quote":"Supplies the self-accelerating cosmological motivation for taking $�lambda_{�rm DGP}$ near the Hubble scale."},{"cited_title":"Nucl. Phys. B 22, 397 (1970); V. I. Zakharov, JETP Lett. 12, 312","cited_arxiv_id":null,"evidence_quote":"Establishes the vDVZ discontinuity that makes the massless-graviton limit subtle."},{"cited_title":"Phys. Lett. B 39, 393","cited_arxiv_id":null,"evidence_quote":"Gives the Vainshtein nonlinear recovery argument that the paper re-examines at $R_*$."},{"cited_title":"Nonperturbative Continuity in Graviton Mass versus Perturbative Discontinuity","cited_arxiv_id":"hep-th/0106001","evidence_quote":"Shows the nonlinear DGP solution approaches general relativity below $R_*$, the result the paper argues lies at the edge of the effective theory."},{"cited_title":"Arkani-Hamed, H","cited_arxiv_id":null,"evidence_quote":"Supplies the massive-gravity Stückelberg and strong-coupling analogy, including the $�Lambda_5$ versus $�Lambda_3$ tuning."},{"cited_title":"No Van Dam-Veltman-Zakharov Discontinuity in Ads Space","cited_arxiv_id":"hep-th/0011152","evidence_quote":"Treats massive gravity on curved backgrounds, the analogue for the curvature-dependent ghost."},{"cited_title":"A Large Mass Hierarchy from a Small Extra Dimension","cited_arxiv_id":"hep-ph/9905221","evidence_quote":"Provides the Randall-Sundrum setup used in the DGP-on-RS section."},{"cited_title":"Nucl. Phys. B 282, 397","cited_arxiv_id":null,"evidence_quote":"Gives the Higuchi ghost result for a light graviton in de Sitter space, the analogue for the $�pi$ ghost."},{"cited_title":"Wave function of the radion in a brane world","cited_arxiv_id":"hep-th/9912160","evidence_quote":"Supplies the radion kinetic term used to identify the IR-brane ghost."}],"review_version":1}