REVIEW 3 major objections 5 minor 24 references
Weak in the boundary: How weak SPT phases spoil anomaly matching
T0 review · 3 major / 5 minor · reviewed 2026-08-06 · deepseek-v4-flash
Pith's one-line read Weak SPT phases on systems with a boundary are not in one-to-one correspondence with weak SPT phases on fully periodic systems; the boundary trivializes a copy of the lower-dimensional weak phases, breaking anomaly inflow.
desk verdict A plausible and new claim that the bulk-boundary map for weak interacting SPTs has kernel equal to the lower-dimensional weak phases, but the key assumption that the boundary is classified by the same cohomology theory is under-justified. read the letter →
The pith
A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.
The reading
What carries the argument
The load-bearing object is the bulk-boundary cohomology map $i^*: D_H^d(T^d) \to D_H^d(T^{d-1})$ induced by including the boundary torus into the bulk torus. The argument combines this map with a standard splitting identity for the torus, $D_H^d(T^{d-1} \times T^1) \cong D_H^d(T^{d-1}) \oplus D_H^{d-1}(T^{d-1})$, which holds for any cohomology theory $D_H$ under the paper's extension of the classification conjecture. Because the bulk torus is $T^{d-1} \times T^1$, the kernel of $i^*$ is forced to be the lower-dimensional summand, making the failure of injectivity a general feature rather than an artifact of one classification scheme.
What would settle it
Take a $2+1$-dimensional weak SPT phase built by stacking one-dimensional $H$-SPT layers in the $d=2$ case, where the predicted kernel is $D_H^1(T^1)$, and put the system on a strip with open boundaries in the stacking direction; if exact diagonalization or a tensor-network calculation finds a protected gapless edge mode or surface anomaly in that geometry, the kernel formula is wrong, while a fully gapped trivial boundary supports it.
Extended reading notes
Core claim
Let $H$ be an internal symmetry and $\mathbb{Z}^d$ the discrete translation symmetry. For fully periodic systems, weak $H$-SPT phases are classified by $D_H^d(T^d)$, the value of the SPT cohomology theory on the real-space unit-cell torus; under the crystalline equivalence principle this is the equivariant cohomology group $H_H^{d+2}(T^d;\mathbb{Z})$. The paper extends the same classification assumption to systems with a boundary, giving $D_H^d(T^{d-1})$ for boundary phases, and defines the bulk-boundary map $i^*$ as the map induced by the inclusion $T^{d-1} \hookrightarrow T^d$. Using a standard splitting identity for the torus, $D_H^d(T^{d-1} \times T^1) \cong D_H^d(T^{d-1}) \oplus D_H^{d-1}(T^{d-1})$, it follows that the kernel of $i^*$ is exactly the second summand. Thus $i^*$ is surjective but not injective: the boundary trivializes a copy of the lower-dimensional weak phases, and the anomaly-inflow interpretation fails for these phases. The same conclusion is illustrated for free fermions in symmetry class AII in $d=3$, where an extra $\mathbb{Z}_2$ weak topological insulator phase is destroyed by the boundary.
Load-bearing premise
The argument assumes that weak SPT phases on a system with a boundary are classified by the same cohomology theory $D_H$ evaluated on the boundary torus $T^{d-1}$; if boundary conditions add extra data or change how the symmetry acts, the bulk-boundary map could be different and the kernel formula would change.
Editorial extensions
If this is right
- For any system with a boundary, the weak SPT classification is $D_H^d(T^{d-1})$, a quotient of the bulk classification by the lower-dimensional group $D_H^{d-1}(T^{d-1})$.
- A nontrivial bulk weak phase lying in the kernel has no gapless edge mode, no surface topological order, and no anomaly to cancel; it is invisible to boundary probes.
- The standard picture that a phase is determined by its bulk alone does not extend to weak phases, because the boundary breaks the translation symmetry that stabilizes them.
- The crystalline equivalence principle must be qualified: either spatial and internal symmetries behave differently at a boundary, or the CEP is not a natural equivalence carrying the bulk-boundary map.
- Strong SPT phases are unaffected: when the torus is collapsed to a point, bulk and boundary classifications coincide.
Reading between the lines
- A testable extension: in a coupled-layer construction of a weak phase built from one-dimensional $H$-SPT layers, opening a boundary in the stacking direction should gap out all edge modes; if protected edge modes survive, the kernel formula needs revision.
- The kernel formula suggests a selection rule for crystalline anomaly matching: the boundary-visible anomaly of a weak phase is determined by its bulk class modulo $D_H^{d-1}(T^{d-1})$, not by the class alone.
- The trivialization by a boundary can be read as a deterministic analogue of disorder-induced trivialization of weak phases, since both break the translation symmetry that defines the phase; this might unify the boundary effect with known fragility of weak topological insulators.
- Since the argument is independent of the crystalline equivalence principle, a direct test in a bosonic lattice model where group cohomology applies should show the same kernel, making the failure observable outside free-fermion systems.
Editorial analysis
A structured set of objections, weighed in public.
Referee Report
Summary. The paper argues that for weak symmetry-protected topological (SPT) phases, the bulk–boundary map is not an isomorphism, contrary to the standard anomaly-inflow picture. Assuming that a generalized cohomology theory D_H classifies SPT phases, the authors posit that weak phases on bulk systems are classified by D_H^d(T^d) and that weak phases on systems with a boundary are classified by the same cohomology theory evaluated on the boundary torus, D_H^d(T^{d-1}). Using the stable James splitting of the torus, they compute the kernel of the natural restriction map i^*: D_H^d(T^d) → D_H^d(T^{d-1}) to be D_H^{d-1}(T^{d-1}) (Eq. 8), meaning a copy of the lower-dimensional weak phases is trivialized by the boundary. The paper concludes that weak SPT phases break anomaly matching and discusses consequences for the crystalline equivalence principle (CEP).
Significance. If the physical identifications made in the paper are accepted, the kernel computation is a clean and uniform result. It applies to any candidate cohomology theory D_H and gives a quantitative statement of why weak SPT phases are not detectable via a boundary, extending the known free-fermion example to interacting systems. The paper is honest about the main postulate, but the abstract and discussion state the result as unconditional. The significance is moderate: the algebraic step is sound, but the physical input that the bulk–boundary map is the pullback in the same cohomology theory is an assumption that needs independent support or clear caveating.
major comments (3)
- [Section IV, Eq. (4)] The identification of the boundary classification as D_H^d(T^{d-1}) with the same cohomology theory D_H is not justified. The equality of strong phases at a point, D_H^d(pt) = \bar D_H^d(pt), does not imply equality of the two cohomology theories on all tori. The sentence “using that weak phases are built from lower dimensional strong phases and D_H and \bar D_H must agree for any d, we see that these cohomology theories must be the same on any torus” is not an argument; it is a leap from pointwise agreement to agreement on all tori. Since the kernel computation in Eq. (8) relies entirely on having the same cohomology theory on both sides, this gap is load-bearing.
- [Section IV, Eqs. (5)–(6)] The physical bulk–boundary map is assumed to be the pullback i^* induced by the embedding T^{d-1} → T^d. This is a nontrivial identification. For free fermions, the standard bulk–boundary map in K-theory is often a boundary homomorphism in a long exact sequence, and the paper does not show that T-duality converts this boundary map into the pullback. The free-fermion discussion in Section III only demonstrates failure of isomorphism; it does not verify that the kernel is given by Eq. (8) for that case. Without a derivation or a precise citation establishing that the physical bulk–boundary map is i^*, the computation of ker i^* does not necessarily describe the phases trivialized by a physical boundary.
- [Abstract and Section VI] The central claim is presented as unconditional (“We show...”, “we have shown above”), but the derivation depends on an extension of Kitaev’s conjecture that the paper itself labels as a postulate in Section IV. The abstract, introduction, and discussion should either state the result as conditional on this postulate or provide independent support for \bar D_H = D_H and for the identification of the physical bulk–boundary map with i^*. As written, the reader is left with a theorem whose main hypothesis is asserted rather than established.
minor comments (5)
- [Section II and Section IV] The degree conventions are inconsistent: Section II classifies d+1-dimensional systems by H^{d+2}_H(T^d; Z), while Section IV uses D_H^d(T^d) for d-dimensional systems. The relation between the two notations should be clarified.
- [Eq. (7)] The James splitting is a stable equivalence of spectra, not an equality of groups; the text should say “isomorphism” and specify whether reduced or unreduced cohomology is being used.
- [Section III] The claim “one can easily check” that the bulk–boundary map is not an isomorphism for free fermions should be supported by an explicit statement of the bulk and boundary K-groups for the cited example, rather than sending the reader to Theorem 3.8 of [5] without commentary.
- [Section V] The phrase “the CEP is not a natural transformation” introduces categorical language without definition or explanation; either define it or rephrase in plain terms.
- [References] Reference [17] contains a typographical error: “X. Liu, , and X. G. Wen” has a stray comma.
Circularity Check
No significant circularity: the boundary-kernel result is a new corollary of an explicit cohomological postulate plus the James splitting, not a re-derivation of its inputs.
full rationale
The paper's central claim, ker i* = D_H^{d-1}(T^{d-1}) in Eq. (8), is reached by combining three ingredients: (1) the explicit postulate that boundary weak SPT phases are classified by some cohomology theory \bar D_H evaluated on T^{d-1}, which is then identified with D_H using strong-phase agreement and the structure of weak phases; (2) the James splitting identity of Eq. (7), which is an independent topological fact; and (3) the embedding-induced pullback i* of Eq. (6). The kernel computation itself is a direct algebraic consequence of these inputs and is not used as an input anywhere. The paper openly labels the boundary classification as an extension of Kitaev's conjecture rather than hiding it as a derived prediction, so the main vulnerability is an unproven assumption about the correct bulk-boundary cohomology theory, not a circular reduction. The self-citations to [20] and [22] are used as prior classification results and as a source for the no-boundary classification and CEP failure; these are independent inputs with content beyond the present boundary-kernel statement. No fitted parameter is renamed as a prediction, and no known result is merely relabeled. Therefore the derivation is self-contained given its stated assumptions, and the appropriate circularity score is 0.
Assumptions & free parameters
assumptions (4)
- domain assumption Kitaev's conjecture: SPT phases are classified by some cohomology theory D_H (e.g., invertible TQFTs, topological gauge theory).
- ad hoc to paper The boundary classification of weak SPT phases is given by the same cohomology theory D_H evaluated on T^{d-1} (i.e., \bar D_H = D_H).
- standard math James splitting / Künneth formula for generalized cohomology of tori: D_H^d(T^{d-1} × T^1) = D_H^d(T^{d-1}) ⊕ D_H^{d-1}(T^{d-1}).
- standard math Free-fermion classification via K-theory with T-duality: KR^{-n}(T^d) = KO_{d-n}(T^d) and boundary version (3).
Cite this review
Pith. "Pith review of Weak in the boundary: How weak SPT phases spoil anomaly matching." pith.science (2026). https://pith.science/paper/D2Z62VC5
@misc{pith2026250717179,
author = {Pith},
title = {Pith review of: Weak in the boundary: How weak SPT phases spoil anomaly matching},
year = {2026},
howpublished = {\url{https://pith.science/paper/D2Z62VC5}},
note = {Machine review of arXiv:2507.17179}
}
read the original abstract
We show how weak symmetry protected topological (SPT) phases on systems with a boundary are not in 1-to-1 correspondence with weak SPT phases on fully periodic systems, breaking the standard anomaly inflow interpretation of SPT phases. We further discuss the implications for the crystalline equivalence principle (CEP).
Reference graph
Works this paper leans on
-
[1]
C. Kane and E. Mele, Quantum spin Hall effect in graphene, Physical Review Letters 95, 226801 (2005)
work page 2005
-
[2]
Fu and C
L. Fu and C. L. Kane, Topological insulators with inver- sion symmetry, Physical Review B—Condensed Matter and Materials Physics 76, 045302 (2007)
2007
- [3]
-
[4]
R. Thorngren and D. V. Else, Gauging spatial sym- metries and the classification of topological crystalline phases, Physical Review X 8, 011040 (2018)
work page 2018
-
[5]
K. Gomi and G. C. Thiang, ‘real’gerbes and dirac cones of topological insulators, Communications in Mathematical Physics 388, 1507 (2021)
work page 2021
-
[6]
E. Prodan and H. Schulz-Baldes, Bulk and Boundary In- variants for Complex Topological Insulators(Springer In- ternational Publishing, 2016)
work page 2016
-
[7]
D. V. Else and C. Nayak, Classifying symmetry-protected topological phases through the anomalous action of the symmetry on the edge, Physical Review B 90, 235137 (2014)
work page 2014
-
[8]
Witten, Fermion path integrals and topological phases, Reviews of Modern Physics 88, 035001 (2016)
E. Witten, Fermion path integrals and topological phases, Reviews of Modern Physics 88, 035001 (2016)
work page 2016
Show all 24 references
-
[9]
D. S. Freed and M. J. Hopkins, Reflection positivity and invertible topological phases, Geometry & Topology 25, 1165 (2021)
2021
-
[10]
Kapustin and R
A. Kapustin and R. Thorngren, Anomalous discrete sym- metries in three dimensions and group cohomology, Phys- 4 ical review letters 112, 231602 (2014)
2014
-
[11]
Cheng, M
M. Cheng, M. Zaletel, M. Barkeshli, A. Vishwanath, and P. Bonderson, Translational symmetry and microscopic constraints on symmetry-enriched topological phases: A view from the surface, Physical Review X 6, 041068 (2016)
2016
-
[12]
C. Wang, A. C. Potter, and T. Senthil, Gapped sym- metry preserving surface state for the electron topologi- cal insulator, Physical Review B—Condensed Matter and Materials Physics 88, 115137 (2013)
2013
-
[13]
X. Chen, F. J. Burnell, A. Vishwanath, and L. Fidkowski, Anomalous symmetry fractionalization and surface topo- logical order, Physical Review X 5, 041013 (2015)
2015
-
[14]
Gomi and G
K. Gomi and G. C. Thiang, Crystallographic t-duality, Journal of Geometry and Physics 139, 50 (2019)
2019
-
[15]
McGreevy, Generalized symmetries in condensed mat- ter, Annual Review of Condensed Matter Physics 14, 57 (2023)
J. McGreevy, Generalized symmetries in condensed mat- ter, Annual Review of Condensed Matter Physics 14, 57 (2023)
2023
-
[16]
Kitaev, Homotopy-theoretic approach to spt phases in action: Z16 classification of three-dimensional supercon- ductors, in Symmetry and Topology in Quantum Matter Workshop (2015)
A. Kitaev, Homotopy-theoretic approach to spt phases in action: Z16 classification of three-dimensional supercon- ductors, in Symmetry and Topology in Quantum Matter Workshop (2015)
2015
-
[17]
X. Chen, Z. C. Gu, X. Liu, , and X. G. Wen, Symmetry protected topological orders and the group cohomology of their symmetry group, Physical Review B 87, 155114 (2013)
2013
-
[18]
Gaiotto and T
D. Gaiotto and T. Johnson-Freyd, Symmetry protected topological phases and generalized cohomology, Journal of High Energy Physics 2019, 7 (2019)
2019
-
[19]
Shiozaki, C
K. Shiozaki, C. Z. Xiong, and K. Gomi, Generalized homology and Atiyah-Hirzebruch spectral sequence in crystalline symmetry protected topological phenomena, arXiv preprint arXiv:1810.00801 (2018)
2018 arXiv
-
[20]
Antol ´ ın-Camarena, A
O. Antol ´ ın-Camarena, A. Debray, C. Krulewski, N. Pacheco-Tallaj, D. Sheinbaum, and L. Stehouwer, Weak topological phases in the presence of interactions, arXiv preprint arXiv:2410.10031 (2024)
2024 arXiv
-
[21]
Hatcher, Algebraic Topology (Cambridge University Press, Cambridge, 2002)
A. Hatcher, Algebraic Topology (Cambridge University Press, Cambridge, 2002)
2002
-
[22]
Sheinbaum and O
D. Sheinbaum and O. Antol ´ ın Camarena, Failure of the crystalline equivalence principle for weak free fermions, Phys. Rev. B 111, L081118 (2025)
2025
-
[23]
Wang, X.-G
J. Wang, X.-G. Wen, and E. Witten, Symmetric gapped interfaces of spt and set states: systematic constructions, Physical Review X 8, 031048 (2018)
2018
-
[24]
Prakash, J
A. Prakash, J. Wang, and T.-C. Wei, Unwinding short- range entanglement, Physical Review B 98, 125108 (2018)
2018
Reviewed August 6, 2026 · model on record in the stance chip above.
Discussion (0). Sign in to comment.