Web of dualities on non-orientable surfaces

Ippo Orii; Keita Tsuji

REVIEW 2 major objections 5 minor 3 references

Web of dualities on non-orientable surfaces

T0 review · 2 major / 5 minor · reviewed 2026-08-03 · deepseek-v4-flash

Challenge this review Re-run · record.json Download PDF Read on arXiv ↗ Review JSON

Pith's one-line read The 16 topological manipulations on time-reversal-symmetric 2d theories form the dihedral group D8.

desk verdict The D8 group claim is very likely right and the SymTFT and sector analysis are genuinely useful, but Section 2.3 has two fixable errors—a false cited ABK identity and mislabeled orbit actions—so the paper needs revision before I'd trust the details. read the letter →

arxiv 2601.09067 v1 pith:L444VUJ6 submitted 2026-01-14 hep-th

Ippo Orii , Keita Tsuji This is my paper · ORCID

classification hep-th

keywords topologicalmanipulationsdihedralgroupD8non-orientablesurfacesPin^-structuresArf-Brown-KervaireinvariantfermionizationSymmetryTFTZ2gauging

verification ladder T0 review T1 audit T2 compute T3 formal

The pith

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The reading

This paper establishes that the topological manipulations available to a two-dimensional bosonic field theory with Z2 and time-reversal symmetry — gauging the Z2, stacking the SPT phase (−1)^{∫A∪w1}, and stacking the Haldane phase (−1)^{∫w2} — together with the corresponding operations on fermionic theories, form the dihedral group D8 of order 16. The authors prove the key relation S_2^B = (O^B S_1^B)^4 by passing to the fermionic description, where the relevant invariant is the Arf–Brown–Kervaire (ABK) invariant of a Pin^- structure, and they prove distinctness of the 16 elements by acting on the trivial fermionic theory, where the ABK invariant produces distinct phases. A reader should care because this turns a seemingly open-ended web of dualities into a finite, classified group of operations: any composition of gauging, fermionization, bosonization, and phase-stacking is one of 16 explicit operations. The paper also re-derives the web in the language of Symmetry TFTs and shows how the operations permute the sectors of the Hilbert space on a circle, using the Majorana/Ising CFTs as a concrete example.

What carries the argument

The load-bearing object is the Z4-valued quadratic refinement q_η associated with a Pin^- structure on a (possibly non-orientable) surface, together with the Arf–Brown–Kervaire invariant ABK(η) = (4/πi) log of the fermionized trivial theory. The identity 4·ABK(η)=∫Σ w2 (mod 8) turns four-fold stacking of the ABK phase into the S_2^B operation, and ABK(η+w1)=−ABK(η) is used to prove that the 16 words act distinctly. The SymTFT description encodes the same manipulations as the exchange of E and M line operators (gauging) and phase actions on M (S_1), with S_2 acting as a global phase on the Hilbert space of the interval.

What would settle it

Directly evaluate the partition function of the trivial fermionic theory on a non-orientable surface with nonzero w2, such as the real projective plane, after applying (O^F S_1^F)^8; the paper predicts the result is identically 1. Any explicit state-sum that yields a different phase would show the group is a proper quotient of D8.

Watch

Extended reading notes

Core claim

The central claim is that the operations O^B (Z2 gauging), S_1^B (stacking (−1)^{∫A∪w1}), and S_2^B (stacking (−1)^{∫w2}) on time-reversal-symmetric 2d bosonic theories, and their fermionic images, satisfy (O^B)^2=(S_1^B)^2=1 and S_2^B=(O^B S_1^B)^4, and that the 16 elements (O^B S_1^B)^n and S_1^B(O^B S_1^B)^n (n=0,...,7) are all distinct. Distinctness is demonstrated on the trivial fermionic theory, whose partition function after the operation (O^F S_1^F)^n is exp(nπi/4 · ABK(η)), with ABK(η) the Z8-valued Arf–Brown–Kervaire invariant; the shift identity ABK(η+w1)=−ABK(η) guarantees the second series of 16 elements is also distinct. Hence the complete group of manipulations is the dihedral

Load-bearing premise

The D8 conclusion depends on imported identities about Arf–Brown–Kervaire invariants — that 4·ABK(η)=∫Σ w2 (mod 8) and ABK(η+w1)=−ABK(η) hold exactly as used — and on restricting to theories whose gravitational anomaly vanishes mod 16 so that bosonization is well-defined.

Editorial extensions

If this is right

Every composition of gauging, fermionization, bosonization, and the two phase-stackings is one of exactly 16 operations with a known action on partition functions, so the web of dualities is finite and classified.
The relation S_2^B=(O^B S_1^B)^4 ties the Haldane phase to four iterations of a gauging-plus-stacking operation, so the 16-element group is generated by only two operations, O^B and S_1^B.
On the S^1 Hilbert space, the operations' action on sectors degenerates to the dihedral group D4 of order 8, with S_2^B acting trivially; the paper works out the sector permutation tables for general theories.
The D8 structure is realized concretely by the Majorana CFT: the sixteen images of the Majorana theory include the Ising CFT (up to the operation O^F S_1^F), and the parity-redefinition freedom in the R sector matches the four-step cycle generated by O^F S_1^F.

Reading between the lines

Editorial extensions of the paper, not claims the author makes directly.

If D8 is correct, the same 16-word group should organize the web for any 2d theory with the same symmetries, including rational CFTs with higher central charge; computing the orbit of each theory under the group would give a new invariant, namely the size of its stabilizer.
The proof that distinctness can be certified solely from the trivial fermionic theory suggests that the ABK invariant itself — not the full partition function — is the universal detector for these manipulations; this may extend to higher dimensions where a similar mod-16 invariant governs bosonization.
The sector degeneracy to D4 hints that the distinction between S_2^B and the identity is invisible in the S^1 Hilbert space and only shows up on non-orientable spacetime surfaces with w2≠0; a direct test on the real projective plane would isolate this.

Share X Bluesky LinkedIn Reddit HN

Signed reviews

No signed human review yet.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, and a circularity audit.

Desk Editor's Note

read the letter

My take: the central claim—that the 2d topological manipulations O^B, S_1^B, S_2^B and their fermionic counterparts generate the dihedral group D8 of order 16—is very likely correct, and the paper's additions are real. The explicit order-16 group identification with the distinctness proof via ABK phases on the trivial theory, the SymTFT realization in §3, and the S^1-sector tables in §4 go beyond what I've seen in [Spi25] and [TY23]. I spot-checked the core algebra: (2.9)–(2.12), the Klein-bottle gauging matrix (4.14), the fermionization matrices (4.34)–(4.35), and the ABK values (4.38) all check out. The Majorana/Ising example is a useful concrete illustration.

But Section 2.3 has two real problems. First, the proof of S_2^B = (O^B S_1^B)^4 invokes the identity 4ABK(η) = ∫ w2. That identity is false as stated. On RP^2, ABK(η)=1 and 4ABK = 4 mod 8, while ∫w2 = 1. What the argument needs is the weaker and true identity ABK(η) ≡ ∫w2 (mod 2). The paper neither states nor proves that. The gap is fixable, but as written the justification for one of the two defining relations of D8 is broken.

Second, the orbit-action display (2.22)–(2.24) has mislabeled indices. Using the paper's own definitions, O^F(T_{F,n}) = T'_{F,n-1}, not T'_{F,-n+1}; S_1^F(T_{F,n}) = T'_{F,n}, not T'_{F,-n}; and O^F S_1^F(T'_{F,n}) = T_{F,n+1}, not T'_{F,n+1}. The paper's (2.20) for Tri_F already contradicts (2.22) at n=0. So Figure 2 needs re-derivation.

Neither issue kills the main theorem. The distinctness argument via ABK phases is sound, and the D8 conclusion should survive once the mod-2 identity is proved and the orbit labels are fixed. But the manuscript as it stands is not fully correct.

Who is this for? People working on 2d dualities, fermionization, non-orientable surfaces, or symmetry TFTs. It deserves a serious referee; with the ABK proof and orbit corrections it would be a solid contribution. I would send it to review.

Referee Report

2 major / 5 minor

Summary. The paper studies a set of topological manipulations on two-dimensional time-reversal-invariant bosonic theories with a non-anomalous Z2 symmetry, and their fermionic images under Pin^- fermionization: gauging (O^B), stacking the SPT (-1)^{∫A∪w1} (S_1^B), stacking the Haldane phase (-1)^{∫w2} (S_2^B), and their fermionic counterparts O^F, S_1^F, S_2^F. The central claim is that these operations generate the dihedral group D8 of order 16, with explicit relations (O^B)^2=(S_1^B)^2=(S_2^B)^2=1 and S_2^B=(O^B S_1^B)^4. The paper also gives a SymTFT interpretation of the operations, analyzes their action on S^1 Hilbert-space sectors (torus and Klein bottle), and illustrates the web with the Majorana/Ising CFT example.

Significance. If the main theorem is correct, the result is a useful and reasonably clean organization of the 2d bosonization/gauging web: the operations are shown to form a finite group with explicit partition-function realizations, including on non-orientable surfaces. The paper contains a substantial amount of explicit algebra, a SymTFT reinterpretation, sector-level tables, and a nontrivial CFT check. The reader verified many of the key formulas, including (2.9)–(2.12), the Klein-bottle gauging matrix (4.14), the fermionization matrices (4.34)–(4.35), and the ABK values (4.38). However, two concrete issues in §2.3 need correction before the proof of the central claim is rigorous.

major comments (2)

[§2.3, Eq. (2.15)] The proof of S_2^B=(O^B S_1^B)^4 uses the statement "4 ABK(η)=∫w2". This equality is false in the paper's own normalization. On Σ=RP^2, with generator a∈H^1, w2=1, and the two Pin^- structures have q_η(a)=1 or 3, so (2.6) gives ABK(η)=1 or 7 mod 8; hence 4 ABK(η)=4 mod 8, not 1. What is actually needed for stacking ABK four times to equal S_2^F is the weaker identity ABK(η) ≡ ∫w2 (mod 2), i.e. exp(πi ABK(η))=(-1)^{∫w2}. This identity appears to be true, but it is neither stated nor proved in the manuscript. Since S_2^B=(O^B S_1^B)^4 is one of the defining relations of D8, the written proof of the main theorem has a genuine gap. Please replace the incorrect statement with the mod-2 relation and supply a proof (for example from the known bordism classification of Pin^- structures or a direct argument).
[§2.3, Eqs. (2.22)–(2.23), Fig. 2] The orbit labels in (2.22)–(2.23) are inconsistent with the explicit fermionic actions (2.12)–(2.14). With T_{F,n}=(O^F S_1^F)^n T_F and T'_{F,n}=S_1^F (O^F S_1^F)^n T_F, direct computation gives O^F T_{F,n} = T'_{F,n-1}, S_1^F T_{F,n} = T'_{F,n}, O^F T'_{F,n} = T_{F,n+1}, and S_1^F T'_{F,n} = T_{F,n} (indices mod 8). For example, (O^F T_0)[η]=T_0[η+w1] exp(πi/4 ABK(η)) equals T'_{F,7}[η], not T'_{F,1}[η]; and (S_1^F T_1)[η]=T_1[η+w1] equals T'_{F,1}[η], not T'_{F,-1}[η]. The signs of n in (2.22)–(2.23) are therefore reversed. This affects the arrows in Figure 2. The separate D8 conclusion (2.17)–(2.18) is not affected, but the generic-orbit description must be corrected.

minor comments (5)

[§2.3, after (2.20)] The sentence "Thus, (O^F S_1^F)^n, S_1^F (O^F S_1^F)^n, n>0 change Tri_F into non-trivial theories" omits n=0, where both operations act trivially on Tri_F; the distinctness of S_1^F from the identity is handled by the following sentence, but the wording should be clarified.
[§2.3, before (2.22)] "In general, each of the 16 elements of D8 maps a fermionic theory T_F to a distinct theory" is not literally true for the trivial theory, since S_1^F(Tri_F)=Tri_F. If "generic" theory is intended, this should be stated.
[§4.2, around (4.39)] The sentence "O^F S_1^F does not affect H_NS. The trace over P=λ sector in H_R of S_1(T_F) is..." seems to have a typo: the second clause presumably refers to O^F S_1^F(T_F), not S_1(T_F). Please rephrase.
[Table 4.2] The table is visually dense and the column headers are ambiguous. Consider splitting the table into separate blocks for the bosonic and fermionic theories, or using clearer subheadings for the untwisted/twisted (H/H_g) and NS/R sectors.
[Footnote 1] The claim that there is "no quadratic-refinement–like object associated with Pin^+ structures" is stated without elaboration. Since this is a side remark, a short explanation or reference would help avoid confusion.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: the D8 group law and 16-element distinctness are derived from the stated definitions plus cited external quadratic-refinement facts, not assumed or fitted.

full rationale

The paper's central claim is that the topological manipulations O^B, S_1^B, S_2^B generate the dihedral group D_8 of order 16. This is not circular. The operations are defined explicitly in (2.1)–(2.3) and their fermionic counterparts are defined by the similarity transformation (2.7). The fermionic expressions (2.12)–(2.14) are derived from these definitions together with standard quadratic-refinement identities. The group relations in (2.15)–(2.16) are then proved, not assumed: S_2^B = (O^B S_1^B)^4 is derived through fermionization and the cited ABK invariant fact, while distinctness of the 16 elements is checked by acting on the trivial fermionic theory and computing partition functions (2.19)–(2.20). That trivial theory is used as an external probe, not as an input that forces the conclusion. The only self-citation is footnote 7, citing the first author's prior work [Ori25] for a Klein-bottle Hilbert-space dimension in an illustrative remark; this is not load-bearing for the D_8 derivation. The imported facts about quadratic refinements and ABK invariants are standard external mathematical results, not self-citations, and are not fitted parameters. The skeptic's concern about the precise normalization of 4ABK(η) = ∫ w2 is a potential mathematical error or gap in the proof as written, not a circularity: the relation is imported, not definitionally identical to the target. Similarly, the noted sign discrepancies in (2.22)–(2.24) would be internal consistency issues, not circularity. Overall, the derivation is self-contained relative to its stated definitions and external invariants, so the circularity score is 0.

Assumptions & free parameters 0 free parameters · 7 assumptions · 0 invented entities

The D8 derivation uses no fitted constants and invokes only standard facts (Pin^- ↔ q_η bijection, ABK identities, Wu formula, Steenrod square) plus the stated non-anomaly and mod-16 anomaly conditions. The SymTFT section adds a modeling choice (trivial T-action on the toric code bulk, fn. 7) that is not needed for the group-theoretic claim. No new particles, forces, or entities are introduced.

assumptions (7)

domain assumption The Z2 and time-reversal symmetries are non-anomalous, so A is valued in H^1(Σ;Z2) on possibly non-orientable Σ.
Footnote 2; the definitions of fermionization (2.4) and gauging (2.1) presuppose this.
standard math Z4-valued quadratic refinements are in one-to-one correspondence with Pin^- structures on closed surfaces, with q_{η+a}(b) = q_η(b) + 2∫a∪b.
Appendix A.2, eqs. (A.27)–(A.28), citing [KT91]; underpins the fermionization formula (2.4).
standard math ABK invariants satisfy 4·ABK(η) = ∫w2 (mod 8) and ABK(η+w1) = −ABK(η) in Z8.
Used in §2.3 to prove S_2^B=(O^B S_1^B)^4 and the order-8 of O^B S_1^B; (A.31) derives the shift law in the paper, the 4ABK=∫w2 relation is cited to [KT91, FK10].
standard math Wu formula: w2 + w1^2 = 0 on any closed 2-manifold, so every 2-manifold admits Pin^- structures.
Appendix A.2, eq. (A.24).
standard math Steenrod square identity A∪A = A∪w1 for A ∈ H^1(Σ;Z2).
Footnote 5, eq. (2.8); used in computing O^F and S_1^F and in the Klein-bottle sector analysis.
domain assumption Gravitational anomaly vanishes mod 16 so that bosonization is equivalent to gauging fermion parity and is well-defined.
Footnote 4; needed for Fer^{-1} and hence for the similarity-transformation definition of the fermionic operations (2.7).
ad hoc to paper In the SymTFT, the bulk toric code is chosen with the parity/time-reversal acting trivially on the e and m lines.
Footnote 7, §3.1; a stated modeling restriction, since non-trivial T-action would break the |e,a⟩-basis (e.g., dim V(K)=2 ≠ |H^1(K;Z2)|=4).

how reviews work

0 comments

Cite this review

Pith. "Pith review of Web of dualities on non-orientable surfaces." pith.science (2026). https://pith.science/paper/L444VUJ6

@misc{pith2026260109067,
  author       = {Pith},
  title        = {Pith review of: Web of dualities on non-orientable surfaces},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/L444VUJ6}},
  note         = {Machine review of arXiv:2601.09067}
}

abstract

It is known that a two-dimensional bosonic theory with a non-anomalous $\mathbb{Z}_2$ symmetry can be fermionized. Recent work shows that if the bosonic theory also has non-anomalous time-reversal symmetry, fermionization extends to non-orientable surfaces and yields a fermionic theory that depends on a $\mathrm{Pin}^-$ structure. Besides fermionization, one can define various topological manipulations, such as gauging and stacking invertible phases, which together generate a web of dualities. We prove that their group structure is the dihedral group $D_8$ of order 16. Furthermore, we systematically investigate the web from two perspectives: Symmetry TFT and actions on sectors of the $S^1$ Hilbert space.

Figures

Figures reproduced from arXiv: 2601.09067 by the authors.

**Figure 2.** illustrates the generic orbit of D8 as a web. 0 0 ′ 1 1 ′ 2 2 ′ 3 3 ′ 4 4 ′ 5 5 ′ 6 6 ′ 7 7 ′ S B 2 S B OB 1 OBS B 1 Fer Fer−1 0 0 ′ 1 1 ′ 2 2 ′ 3 3 ′ 4 4 ′ 5 5 ′ 6 6 ′ 7 7 ′ S F 2 S F OF 1 OF S F 1 [PITH_FULL_IMAGE:figures/full_fig_p008_2.png] view at source ↗

**Figure 3.** Symmetry TFT provides a framework for understanding the partition functions [PITH_FULL_IMAGE:figures/full_fig_p011_3.png] view at source ↗

Figures from the paper (2 more)

**Figure 4.** Figure 4: An operation on the theory is described as attaching a domain wall from the bulk [PITH_FULL_IMAGE:figures/full_fig_p011_4.png]

**Figure 5.** Figure 5: The torus T 2 is represented as a cylinder with its ends identified, with time running upward. The Poincar´e duals of the spatial and temporal cycles define x, y ∈ H1 (T 2 ; Z2). For the Klein bottle, an orientation-reversing defect is inserted along the spatial direct…

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

3 extracted references · 3 linked inside Pith

[1]

Apruzzi, F

[ABGE+21] F. Apruzzi, F. Bonetti, I. Garc ´ ıa Etxebarria, S. S. Hosseini, and S. Schafer- Nameki,Symmetry TFTs from String Theory, Commun. Math. Phys.402 (2023) 895–949, arXiv:2112.02092 [hep-th]. [BBJ+16] M. Barkeshli, P. Bonderson, C.-M. Jian, M. Cheng, and K. Walker,Reflection and time reversal symmetry enriched topological phases of matter: path inte...

arXiv 2023
[1997]

[DJL23] Z. Duan, Q. Jia, and S. Lee,Z N duality and parafermions revisited, JHEP11 (2023) 206, arXiv:2309.01913 [hep-th]. [FK10] L. Fidkowski and A. Kitaev,Topological phases of fermions in one dimension, Phys. Rev. B83(2011) 075103, arXiv:1008.4138 [cond-mat.str-el]. [FMT22] D. S. Freed, G. W. Moore, and C. Teleman,Topological symmetry in quantum field t...

arXiv 2023
[2023]

[SN23] S

arXiv:2308.00747 [hep- th]. [SN23] S. Schafer-Nameki,ICTP lectures on (non-)invertible generalized symmetries, Phys. Rept.1063(2024) 1–55, arXiv:2305.18296 [hep-th]. [Spi25] R. C. Spieler,Exploring two dimensionalZ 2 invariant phases with time reversal symmetry and their transitions with topological operations, JHEP10(2025) 069, arXiv:2504.20021 [cond-mat...

arXiv 2024

Pith tools

Reviewed August 3, 2026 · model on record in the stance chip above.

[1] [1]

Apruzzi, F

[ABGE+21] F. Apruzzi, F. Bonetti, I. Garc ´ ıa Etxebarria, S. S. Hosseini, and S. Schafer- Nameki,Symmetry TFTs from String Theory, Commun. Math. Phys.402 (2023) 895–949, arXiv:2112.02092 [hep-th]. [BBJ+16] M. Barkeshli, P. Bonderson, C.-M. Jian, M. Cheng, and K. Walker,Reflection and time reversal symmetry enriched topological phases of matter: path inte...

arXiv 2023

[2] [1997]

[DJL23] Z. Duan, Q. Jia, and S. Lee,Z N duality and parafermions revisited, JHEP11 (2023) 206, arXiv:2309.01913 [hep-th]. [FK10] L. Fidkowski and A. Kitaev,Topological phases of fermions in one dimension, Phys. Rev. B83(2011) 075103, arXiv:1008.4138 [cond-mat.str-el]. [FMT22] D. S. Freed, G. W. Moore, and C. Teleman,Topological symmetry in quantum field t...

arXiv 2023

[3] [2023]

[SN23] S

arXiv:2308.00747 [hep- th]. [SN23] S. Schafer-Nameki,ICTP lectures on (non-)invertible generalized symmetries, Phys. Rept.1063(2024) 1–55, arXiv:2305.18296 [hep-th]. [Spi25] R. C. Spieler,Exploring two dimensionalZ 2 invariant phases with time reversal symmetry and their transitions with topological operations, JHEP10(2025) 069, arXiv:2504.20021 [cond-mat...

arXiv 2024