Equivalence of different definitions of higher $\rho$ invariants

Guoliang Yu; Hongzhi Liu; Zhizhang Xie

REVIEW 3 major objections 6 minor 21 references

Equivalence of different definitions of higher $\rho$ invariants

T0 review · 3 major / 6 minor · reviewed 2026-08-12 · deepseek-v4-flash

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

Pith's one-line read This paper proves that the two standard definitions of the higher rho invariant agree for orientation-preserving homotopy equivalences.

desk verdict Solid even-dimensional proof of a known-but-unwritten equivalence; odd-dimensional case is a sketch and the abstract overstates it. read the letter →

arxiv 2411.16716 v1 pith:RV2TCKX2 submitted 2024-11-22 math.AT math.OA

Hongzhi Liu , Zhizhang Xie , Guoliang Yu This is my paper · ORCID

classification math.ATmath.OA MSC 19K5658J2246L8057R67

keywords higherrhoinvarianthomotopyequivalencesignatureoperatorHilbertPoincarécomplexRoealgebralocalizedindexK-theorysecondary

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

The paper proves that the two standard ways of defining the higher rho invariant attached to an orientation-preserving homotopy equivalence between closed oriented smooth manifolds produce the same invariant. The Piazza-Schick construction uses a bounded replacement of the pullback map on differential forms, obtained through the Hilsum-Skandalis embedding, and lands in the K-theory of a structure algebra. The Higson-Roe construction works with analytically controlled Hilbert Poincaré complexes and has both a piecewise-linear and a smooth version. The paper shows the two Higson-Roe versions agree with each other, and that the smooth version coincides with the Piazza-Schick invariant through the localized index isomorphism. The result matters because the higher rho invariant is the secondary invariant that detects whether a homotopy equivalence can be deformed into a homeomorphism.

What carries the argument

The central object is the $\Gamma$-equivariant analytically controlled Hilbert Poincaré complex, a chain complex of Hilbert spaces over the universal cover equipped with a self-adjoint Poincaré duality operator $S$ whose differential $D=d+d^*$ satisfies controlled conditions placing its resolvents in the Roe algebra. Its signature class is the formal difference $[P_+(D+S)]-[P_+(D-S)]$ in $K_0(C^*(\widetilde N)^\Gamma)$. The proof is carried by Higson and Roe's path-of-projections construction, which shows this signature class is invariant under analytically controlled chain homotopy equivalence, together with Roe's localized index formula, which transports $\rho_{\mathrm{op}}(f)$ into the same $K_0$ group. Explicit paths $P[A](t)-Q[A](t)$ of differences of projections interpolate between the representatives of $\rho_{\mathrm{PL}}(f)$, $\rho_{C^\infty}(f)$, and $\mathrm{Ind}_L[\rho_{\mathrm{op}}(f)]$.

What would settle it

Take any orientation-preserving homotopy equivalence $f\colon M\to N$ with nontrivial fundamental group whose class in the topological structure group is nonzero, and compute $\mathrm{Ind}_L[\rho_{\mathrm{op}}(f)]$ and $[\rho_{C^\infty}(f)]$ in $K_0(C^*_{L,0}(\widetilde N)^\Gamma)$; a difference between the two classes would refute the main theorem. A more local check is whether the path of projections in Lemma 5.2 fails to lie in the localization algebra for some $f$.

Watch

Extended reading notes

Core claim

Let $f\colon M\to N$ be an orientation-preserving homotopy equivalence of closed oriented smooth manifolds with common fundamental group $\Gamma$, and let $\widetilde M$ and $\widetilde N$ be the universal covers. The paper's central claim is that the Piazza-Schick higher rho invariant $\rho_{\mathrm{op}}(f)$ and the Higson-Roe higher rho invariants $\rho_{\mathrm{PL}}(f)$ and $\rho_{C^\infty}(f)$ all represent the same secondary class. The two main results are Theorem 5.1: $[\rho_{\mathrm{PL}}(f)]=[\rho_{C^\infty}(f)]$ in $K_0(C^*_L(\widetilde N)^\Gamma)$, and Theorem 5.6: $\mathrm{Ind}_L[\rho_{\mathrm{op}}(f)]=[\rho_{C^\infty}(f)]$ in $K_0(C^*_L(\widetilde N)^\Gamma)$. Since $\mathrm{Ind}_L$ is an isomorphism from $K_1(D^*(\widetilde N)^\Gamma)$ to $K_0(C^*_{L,0}(\widetilde N)^\Gamma)$, the two definitions of the higher rho invariant are equivalent.

Load-bearing premise

The proof needs the Hilsum-Skandalis construction to replace the pullback map on differential forms with a bounded operator that behaves like a chain homotopy equivalence; if that construction fails for some homotopy equivalence, the Piazza-Schick invariant is undefined and the equivalence cannot hold.

Editorial extensions

If this is right

A single higher rho invariant class in $K_0(C^*_{L,0}(\widetilde N)^\Gamma)$ now describes all three constructions, so the invariant is framework-independent.
If any of the equivalent representatives is nonzero, the orientation-preserving homotopy equivalence $f$ cannot be deformed into a homeomorphism.
The equality $[\rho_{\mathrm{PL}}(f)]=[\rho_{C^\infty}(f)]$ shows that the invariant does not depend on whether the manifold is described by a PL triangulation or by smooth differential forms.
The explicit interpolating paths give a recipe for transferring computations between the bounded (simplicial) and unbounded (de Rham) settings.

Reading between the lines

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

This suggests the same path-of-projections comparison could define higher rho invariants for other classes of manifolds, such as topological or Witt spaces, whenever a bounded replacement map like $T_{\widetilde f}$ is available.
The equivalence also indicates that the Hilsum-Skandalis embedding is the single external ingredient on which the Piazza-Schick definition depends; a different or simplified embedding would automatically yield the same unified invariant.
One could use the equivalence to compute the invariant in concrete examples using whichever side is easier, for example the de Rham version for analytic index estimates and the PL version for combinatorial surgery-theoretic arguments.

Share X Bluesky LinkedIn Reddit HN

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

Quick take: the paper does what the title says, with one real caveat—the proof is written out for even-dimensional manifolds, and the odd-dimensional case is disposed of with a sketch. The abstract claims all closed oriented smooth manifolds, which is not what is actually proved. Still, the core equivalence is real and the paper is a useful service to the area.

What's new: the comparison between the Piazza–Schick (Hilsum–Skandalis) higher rho and the Higson–Roe/Weinberger–Xie–Yu simplicial higher rho has been folklore, and Jiang–Liu looked at the de Rham version, but this is the first full proof that they land on the same K-theory class. The route is sensible: show the bounded and unbounded Higson–Roe versions agree, then identify the unbounded version with the Hilsum–Skandalis construction. The even-dimensional chain of equalities in Theorems 5.1 and 5.6 is laid out carefully, and the paper is honest about using external results (Hilsum–Skandalis embedding, Wahl's controlled chain homotopy equivalence, Higson–Roe's signature invariance). No circularity: the invariants being compared are defined independently.

Soft spots, in order of size. First, the odd-dimensional case. The introduction waves it away with 'as usual' or 'invert 2 and cross with S1'. Neither is a proof. Crossing with S1 requires naturality of rho_op and rho_C∞ under suspension, including the behavior of the Hilsum–Skandalis operator under product with S1, and commutativity of Ind_L with Bott periodicity. None of that appears. This is almost certainly fixable, but as written the main theorem is proved for even dimensions. Second, Lemma 5.4 is the pivot of the proof of Theorem 5.6, and it is imported from Wahl rather than proved or precisely stated here. Acceptable if the reference is solid, but it makes the paper less self-contained than the intro promises. Minor: a few display typos in the path definitions in Corollary 5.5, and some notation overload in Lemma 5.2. Cosmetic, not substantive.

Citations look fair. The reliance on [1], [17], [18] is appropriate given they proved the ingredients. No invented entities, no free parameters.

Bottom line: a solid, useful comparison paper. It deserves proper peer review. The referee should push for a real odd-dimensional argument or a corrected statement, and a polishing pass. I would bring it to our reading group and cite it if I worked on higher rho invariants.

Referee Report

3 major / 6 minor

Summary. The paper addresses the equivalence of two existing definitions of the higher rho invariant associated to an orientation-preserving homotopy equivalence f: M -> N of closed oriented smooth manifolds: the Piazza-Schick invariant rho_op, based on the Hilsum-Skandalis perturbation of the signature operator, and the Higson-Roe invariant, based on analytically controlled Hilbert Poincare complexes. The strategy is to prove that the bounded and unbounded versions of the Higson-Roe approach agree, and then that the unbounded version agrees with the Hilsum-Skandalis approach. Concretely, the paper proves Theorem 5.1, that rho_PL(f) equals rho_Cinfty(f) in K-theory, and Theorem 5.6, that Ind_L(rho_op(f)) equals rho_Cinfty(f), where Ind_L is the localized index isomorphism from K_1(D*(N~)^Gamma) to K_0(C^*_{L,0}(N~)^Gamma). The paper states in the introduction that the odd-dimensional case is either completely parallel or reduces to the even-dimensional case by taking a product with a circle after inverting 2, and it then restricts attention to even-dimensional manifolds throughout Sections 4 and 5.

Significance. If the stated equivalence is established in full generality, this is a valuable unification of two important secondary invariants in noncommutative geometry and surgery theory. The paper gives a fairly explicit chain of equalities: Lemma 4.1 identifies the simplicial and de Rham K-homology classes of the signature operator; Theorem 5.1 identifies rho_PL and rho_Cinfty; Theorem 5.6 identifies Ind_L(rho_op) with rho_Cinfty. The proofs use established results from Higson-Roe, Hilsum-Skandalis, Wahl, and Weinberger-Xie-Yu, and the comparison is not circular because the two invariants are defined independently. The main weakness is that the odd-dimensional case, which is part of the abstract's claim, is only asserted and not proved, and a few key steps in the final comparison are too compressed.

major comments (3)

[Section 1] The abstract and title claim equivalence for all orientation-preserving homotopy equivalences of closed oriented smooth manifolds, but all proofs in Sections 4 and 5 are carried out only in even dimensions. For example, Section 4.1 begins 'For an n = 2k dimensional manifold', Section 4.2 uses chiral duality with exponent n/2, and Definition 3.8(2) defines the signature class only for even-dimensional Hilbert Poincare complexes. The statement in the introduction that the odd-dimensional case is 'completely parallel' or reduces to the even case by taking product with S^1 after inverting 2 is not a proof: one would need to show naturality of rho_op, rho_Cinfty, and Ind_L under suspension, and to track the Hilsum-Skandalis operator T_{(f x S^1)~} through the tensor-product decomposition L^2(Lambda*(N x S^1)) congruent to L^2(Lambda*(N)) tensor L^2(Lambda*(S^1)). That is a non-formal operator-level computation that the paper nowhere records. Since the odd-dimensional case is part of the stated theorem, this is a load-bearing gap.
[Section 5.5, proof of Theorem 5.6] The final step of the proof contains the key identification: 'For alpha = beta = 0, one can directly see that the representative elements in Corollary 5.5 represents the class [rho_Cinfty(f)] - [rho_Cinfty(I)].' This is asserted without demonstration. The paths displayed in Corollary 5.5 are built from S_{alpha,beta,f sqcup I}(t) and D_{alpha,beta}, whereas Definition 5.3 of rho_Cinfty is built from P[T_f~](3-t) - Q[T_f~](3-t) and F_Cinfty. The equality of these two paths in K_0(C^*_{L,0}(N~)^Gamma) is precisely the final comparison being proved, so it requires an explicit argument rather than a 'directly see' assertion.
[Section 5.5, Lemma 5.2] Lemma 5.2 is the bridge between Roe's index formula for Ind_L(rho_op(f)) and the path involving F_Cinfty, but its proof is not complete. The operator W(t) is introduced as V(t)G_{alpha,beta}(t), where V(t) is a square root of S_{alpha,beta}(t)G_{alpha,beta}(t), and the proof uses conjugation by W(t) to identify two formal differences of projections. The text does not verify that W(t) and V(t) belong to the appropriate localization or structure algebras, nor that conjugation by W(t) preserves the class in K_0(C^*_{L,0}(N~)^Gamma). Since the lemma is load-bearing for the main theorem, these details need to be provided.

minor comments (6)

[Section 1] The first sentence contains a typo: 'Fredhoml' should be 'Fredholm'.
[Section 3 heading] The heading contains a typo: 'Poicnar\'e' should be 'Poincar\'e'.
[Theorems 5.1 and 5.6] The statements say the equalities hold in K_0(C^*_L(N~)^Gamma), but the definitions in Sections 5.2 and 5.3 place rho_PL(f) and rho_Cinfty(f) in K_0(C^*_{L,0}(N~)^Gamma), and Ind_L takes values in K_0(C^*_{L,0}(N~)^Gamma); the target K-group should be corrected consistently.
[Section 5.2] The sentence defining [rho_PL(f)] says the class lies in K_0(C^*_{L,0}(~M)^Gamma), but the path is constructed over ~N; it should be K_0(C^*_{L,0}(~N)^Gamma).
[References] Reference [5] is listed without a year or publication status; if it is a preprint, that should be stated.
[Sections 5.2 and 5.3] The interval notation 't in [3, infinity]' should be '[3, infinity)' in the definitions of rho_PL(f) and rho_Cinfty(f).

Circularity Check

0 steps flagged · score 2.0 of 10

No circular reduction found; the equivalence proof compares independently defined invariants via explicit operator paths, with only non-circular self-citations.

full rationale

The paper's central claim is an equivalence between two invariants defined in different frameworks: rho_op follows Piazza–Schick/Hilsum–Skandalis, while rho_PL and rho_C-infinity follow Higson–Roe and Weinberger–Xie–Yu. The proof proceeds by constructing explicit paths of projections and invoking homotopy invariance of signature classes, rather than by identifying the two definitions by construction. Lemma 5.2 computes Ind_L[rho_op(f)] via Roe's index formula and Lemmas 4.2–4.3, which are independent operator computations. Theorem 5.6 then connects this representative to rho_C-infinity by a further path argument. The use of self-citations, such as Lemma 5.3 citing [1, Theorem 5.2] for triviality of the identity map, is load-bearing but not circular: the cited theorem does not assume the equivalence being proved and concerns a different statement. Other self-citations, e.g. [17] and [18], supply definitions and standard isomorphisms, not the target equality. The main unaddressed issue is the odd-dimensional case, which the introduction dispatches by 'without loss of generality' and a claimed suspension reduction without proof; this is a completeness/correctness gap, not a circularity. No parameter fitting, renaming, or definitional identification forces the result, so the derivation chain is not circular.

Assumptions & free parameters 0 free parameters · 5 assumptions · 0 invented entities

The paper introduces no fitted parameters. The central result rests on prior theorems about Hilsum-Skandalis perturbations, Higson-Roe homotopy invariance of signature classes, the localization index isomorphism, and bounded-geometry subdivisions. The most fragile input is the unproved odd-to-even reduction, which is asserted in the Introduction. The proof is otherwise an assembly of existing machinery.

assumptions (5)

domain assumption The pullback map f~ induces a bounded, analytically controlled chain homotopy equivalence T_f~ via the Hilsum-Skandalis embedding.
Invoked in Section 5.1 to construct rho_op; cited from Hilsum-Skandalis [6] and Wahl [16], not proved in this paper.
standard math The signature class of an analytically controlled Hilbert Poincare complex is invariant under controlled chain homotopy equivalence.
Used in Sections 3.3 and 4 to build paths of projections; due to Higson-Roe [2, Lemma 5.7 and 5.8].
standard math The localization index map Ind_L: K_1(D^*(N~)^Γ) -> K_0(C^*_{L,0}(N~)^Γ) is an isomorphism, and K_*(C^*_{L,0}(N~)^Γ) is isomorphic to K_{*+1}(D^*(N~)^Γ).
Used in Section 5 to pass between the Piazza-Schick and Higson-Roe receptacles; cited from Xie-Yu [18, Section 6] and Yu [21].
domain assumption For every closed manifold N~, a Γ-invariant triangulation admits uniformly bounded-geometry subdivisions Sub_n(N~), and the simplicial and de Rham complexes are controlled Hilbert Poincare complexes.
Used in Section 4.1 to define the paths F_PL; cited from Weinberger-Xie-Yu [17, Section 4.2] and Higson-Xie [5].
ad hoc to paper The odd-dimensional case reduces to the even-dimensional case by taking product with the circle after inverting 2.
Asserted in the Introduction: 'if we invert 2, the odd dimensional case reduces to the even dimensional case by taking direct product with the circle'. The reduction is not proved or cited.

how reviews work

0 comments

Cite this review

Pith. "Pith review of Equivalence of different definitions of higher $\rho$ invariants." pith.science (2026). https://pith.science/paper/RV2TCKX2

@misc{pith2026241116716,
  author       = {Pith},
  title        = {Pith review of: Equivalence of different definitions of higher $\rho$ invariants},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/RV2TCKX2}},
  note         = {Machine review of arXiv:2411.16716}
}

abstract

For each orientation-preserving homotopy equivalence between two closed oriented smooth manifolds, there are mainly two different approaches to the higher $\rho$ invariant associated to this homotopy equivalence. In this article, we show that these two definitions of the higher $\rho$ invariant are equivalent.

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

21 extracted references · 21 canonical work pages

[1]

X. Chen, H. Liu, and G. Yu. Higher ρ invariant is an obstruction to the inverse being local. J. Geom. Phys. , 150:103592, 18, 2020

work page 2020
[17]

Weinberger, Z

S. Weinberger, Z. Xie, and G. Yu. Additivity of higher rho invarian ts and nonrigidity of topological manifolds. Comm. Pure Appl. Math. , 74(1):3–113, 2021

work page 2021
[18]

Xie and G

Z. Xie and G. Yu. Positive scalar curvature, higher rho invariant s and local- ization algebras. Adv. Math., 262:823–866, 2014

work page 2014
[2]

Higson and J

N. Higson and J. Roe. Mapping surgery to analysis. I. Analytic sign atures. K-Theory, 33:277–299, 2005

work page 2005
[3]

Higson and J

N. Higson and J. Roe. Mapping surgery to analysis. II. Geometric signatures. K-Theory, 33:301–324, 2005

work page 2005
[4]

Higson and J

N. Higson and J. Roe. Mapping surgery to analysis. III. Exact se quences. K- Theory, 33:325–346, 2005

work page 2005
[5]

Higson and Z

N. Higson and Z. Xie. Higher signatures of Witt spaces. preprint. EQUIV ALENCE OF HIGHER ρ INV ARIANTS 27

work page
[6]

Hilsum and G

M. Hilsum and G. Skandalis. Invariance par homotopie de la signatur e ` a coef- ﬁcients dans un ﬁbr´ e presque plat. J. Reine Angew. Math. , 423:73–99, 1992

work page 1992

Show all 21 references

[7]

Jiang and H

B. Jiang and H. Liu. Additivity of higher rho invariant for the topolo gical struc- ture group from a diﬀerential point of view. J. Noncommut. Geom., 14(2):441– 486, 2020

2020
[8]

G. G. Kasparov. Operator K-theory and its applications: elliptic operators, group representations, higher signatures, C∗ -extensions. In Proceedings of the International Congress of Mathematicians, Vol. 1, 2 (Warsa w, 1983) , pages 987–1000. PWN, Warsaw, 1984

1983
[9]

A. S. Mishchenko. Homotopy invariants of nonsimply connected m anifolds. I. Rational invariants. Izv. Akad. Nauk SSSR Ser. Mat. , 34:501–514, 1970

1970
[10]

A. S. Mishchenko. Inﬁnite-dimensional representations of dis crete groups, and higher signatures. Izv. Akad. Nauk SSSR Ser. Mat. , 38:81–106, 1974

1974
[11]

S. P. Novikov. Algebraic construction and properties of Hermit ian analogs of K-theory over rings with involution from the viewpoint of Hamiltonian formalism. Applications to diﬀerential topology and the theory of ch aracteristic classes. I. II. Izv. Akad. Nauk SSSR Ser. Mat. ...

1970
[12]

Piazza and T

P. Piazza and T. Schick. Bordism, rho-invariants and the Baum- Connes con- jecture. J. Noncommut. Geom. , 1(1):27–111, 2007

2007
[13]

Piazza and T

P. Piazza and T. Schick. The surgery exact sequence, K-theo ry and the signa- ture operator. Ann. K-Theory, 1(2):109–154, 2016

2016
[14]

J. Roe. Coarse cohomology and index theory on complete Rieman nian mani- folds. Mem. Amer. Math. Soc. , 104(497):x+90, 1993

1993
[15]

J. Roe. Index theory, coarse geometry, and topology of manifolds , volume 90 of CBMS Regional Conference Series in Mathematics . Published for the Confer- ence Board of the Mathematical Sciences, Washington, DC; by the American Mathematical Society, Providence, RI, 1996

1996
[16]

C. Wahl. Higher ρ-invariants and the surgery structure set. J. Topol., 6(1):154– 192, 2013

2013
[19]

Xie and G

Z. Xie and G. Yu. Higher rho invariants and the moduli space of po sitive scalar curvature metrics. Adv. Math. , 307:1046–1069, 2017

2017
[20]

Xie and G

Z. Xie and G. Yu. Higher invariants in noncommutative geometry. In Advances in noncommutative geometry , pages 691–751. Springer, Cham, [2019] ©2019

2019
[21]

G. Yu. Localization algebras and the coarse Baum-Connes conj ecture. K- Theory, 11(4):307–318, 1997. 28 H. LIU, Z. XIE, AND G. YU H. Liu, School of Mathematics, Shanghai University of Finan ce and Economics, Shanghai, 200433, P. R. China. e-mail: liu.hong zhi@sufe.edu.cn Z. Xi...

1997

Pith tools

Reviewed August 12, 2026 · model on record in the stance chip above.

[1] [1]

X. Chen, H. Liu, and G. Yu. Higher ρ invariant is an obstruction to the inverse being local. J. Geom. Phys. , 150:103592, 18, 2020

work page 2020

[2] [17]

Weinberger, Z

S. Weinberger, Z. Xie, and G. Yu. Additivity of higher rho invarian ts and nonrigidity of topological manifolds. Comm. Pure Appl. Math. , 74(1):3–113, 2021

work page 2021

[3] [18]

Xie and G

Z. Xie and G. Yu. Positive scalar curvature, higher rho invariant s and local- ization algebras. Adv. Math., 262:823–866, 2014

work page 2014

[4] [2]

Higson and J

N. Higson and J. Roe. Mapping surgery to analysis. I. Analytic sign atures. K-Theory, 33:277–299, 2005

work page 2005

[5] [3]

Higson and J

N. Higson and J. Roe. Mapping surgery to analysis. II. Geometric signatures. K-Theory, 33:301–324, 2005

work page 2005

[6] [4]

Higson and J

N. Higson and J. Roe. Mapping surgery to analysis. III. Exact se quences. K- Theory, 33:325–346, 2005

work page 2005

[7] [5]

Higson and Z

N. Higson and Z. Xie. Higher signatures of Witt spaces. preprint. EQUIV ALENCE OF HIGHER ρ INV ARIANTS 27

work page

[8] [6]

Hilsum and G

M. Hilsum and G. Skandalis. Invariance par homotopie de la signatur e ` a coef- ﬁcients dans un ﬁbr´ e presque plat. J. Reine Angew. Math. , 423:73–99, 1992

work page 1992