---
title: Wall-Chebyshev Projector
url: https://www.emergentmind.com/topics/wall-chebyshev-projector
type: topic
---

# Wall-Chebyshev Projector

The **wall-Chebyshev projector** is a polynomial spectral projector obtained by approximating an idealized “wall” function with Chebyshev polynomials after an affine rescaling of the Hamiltonian spectrum. In the many-body formulations summarized in the literature, it is the infinite-imaginary-time limit of the exponential projector and is designed to retain the ground-state component while suppressing all excited-state components. The construction appears in deterministic configuration-interaction projector methods, in stochastic projector-based quantum Monte Carlo, and in quantum algorithms for ground-state preparation; closely related Chebyshev polynomial projectors also arise in interval filtering for singular-value problems and in occupied-subspace projection for Kohn–Sham density functional theory [1606.08475; 2402.16685; 2508.00533].

## 1. Definition and spectral normalization

In imaginary-time propagation, one starts from
\[
\Psi(\beta)=e^{-\beta(\hat H-S)}\Psi(0),
\]
and, with a shift chosen such that \(E_0\le S<E_1\), the limit
\[
\hat W=\lim_{\beta\to\infty}e^{-\beta(\hat H-S)}
\]
annihilates all components with \(E>E_0\) while leaving the ground-state component intact. For any trial state with nonzero overlap onto the true ground state, \(\hat W\Psi(0)\propto\Psi_0\) [2402.16685].

The projector is expressed in polynomial form after mapping the Hamiltonian spectrum to \([-1,1]\). One common rescaling is
\[
\tilde H=\frac{2\hat H-(E_{N-1}+E_0)}{E_{N-1}-E_0},
\]
while a closely related form used in quantum algorithms is
\[
\tilde H = 2\,(H-E_0)/R -1,\qquad R=E_{\max}-E_0.
\]
Under these maps, the ground-state eigenvalue is placed at the endpoint or discontinuity of the wall function, and the remaining spectrum is placed in the complementary region [2402.16685; 2508.00533].

The wall function itself is written in slightly different but equivalent normalized forms, depending on the spectral map. In one convention,
\[
\mathrm{wall}(x)=
\begin{cases}
\infty,&x<0,\\
1,&x=0,\\
0,&x>0,
\end{cases}
\]
and in another,
\[
W(x)=
\begin{cases}
\infty,&x\le -1,\\
1,&x=-1,\\
0,&x>-1.
\end{cases}
\]
This suggests that the defining feature is not the particular location of the discontinuity, but the use of an endpoint-localized step-like function that singles out the ground-state eigenvalue after rescaling [2402.16685; 2508.00533].

## 2. Chebyshev-series construction

The construction employs Chebyshev polynomials of the first kind,
\[
T_0(x)=1,\qquad T_1(x)=x,\qquad T_{k+1}(x)=2x\,T_k(x)-T_{k-1}(x),
\]
or equivalently \(T_k(\cos\theta)=\cos(k\theta)\). Because these polynomials are orthogonal on \([-1,1]\) with weight \((1-x^2)^{-1/2}\), they furnish near-minimax polynomial approximations for analytic functions on that interval [1606.08475].

In the deterministic projector formulation, the wall-Chebyshev generator is derived as the \(\tau\to\infty\) limit of an \(m\)th-degree Chebyshev approximation to the exponential propagator. With
\[
R=\tfrac12(E_{N-1}-E_0),\qquad \xi=\frac{x-E_0-R}{R},
\]
the finite-\(\tau\) exponential-Chebyshev approximation is
\[
g_m(x)=C_m(\tau R)\sum_{k=0}^m (2-\delta_{k0})\,I_k(\tau R)\,T_k(-\xi),
\]
and using \(\lim_{\tau\to\infty}I_{k+1}(\tau R)/I_k(\tau R)=1\), one obtains the wall limit
\[
g_m^{\rm wall}(x)=\frac{1}{2m+1}\sum_{k=0}^m (2-\delta_{k0})\,T_k(-\xi).
\]
In the normalized endpoint formulation, the same projector appears as
\[
g^{\rm wall\!-\!Ch}_m(\hat H)
=\frac{1}{1+2m}\sum_{k=0}^m (2-\delta_{k0})(-1)^k\,T_k(\tilde H).
\]
The coefficients are therefore explicit and require no numerical fitting [1606.08475; 2402.16685].

The literature assigns two distinct kinds of performance statements to this construction. In projector analysis, the wall-Chebyshev generator is described as **optimal** in the sense of maximizing the asymptotic convergence factor
\[
\gamma\approx \frac{m(m+1)}{3R}
\]
for a given polynomial degree \(m\) [1606.08475]. In stochastic QMC analysis, the uniform approximation error at the first excitation energy is reported to scale asymptotically as \(1/m\), so achieving \(\|g_m(E_1)\|\le\varepsilon\) requires \(m=O(1/\varepsilon)\) [2402.16685]. These statements concern different quantities: one addresses asymptotic ground-state filtering, the other pointwise approximation of the wall function.

## 3. Deterministic projector configuration interaction

A deterministic many-body realization appears in projector configuration interaction (PCI), which combines projection onto the ground state with a path-filtering truncation scheme and is formulated as a deterministic version of full configuration interaction quantum Monte Carlo (FCIQMC). In this setting, one works in a determinant basis \(\{\lvert\Phi_I\rangle\}\) and propagates
\[
\lvert\Omega^{(n)}\rangle=\sum_J C_J^{(n)}\lvert\Phi_J\rangle.
\]
For a linearized projector step, the spawning amplitudes are
\[
A_{IJ}^{(n+1)}
=\frac{1}{\tau}\,
\bigl\langle\Phi_I\bigr|
\bigl[1-\tau(\hat H-E_0)\bigr]
\bigl|\Phi_J\bigr\rangle
\,C_J^{(n)},
\]
and path filtering imposes a single spawning-threshold \(\varepsilon\) through
\[
A_{IJ}^{(n+1)}(\varepsilon)
=
A_{IJ}^{(n+1)}
\,\Theta\!\bigl(|A_{IJ}^{(n+1)}|-\varepsilon\bigr),\qquad I\neq J.
\]
Diagonal amplitudes are kept exact, coefficients are updated from the retained amplitudes, and the state is renormalized [1606.08475].

Within this framework, the wall-Chebyshev projector replaces the linearized imaginary-time propagator by the infinite-step optimal Chebyshev generator. The path-filtering threshold is the sole truncation parameter, and the resulting error in energy is reported to scale roughly linearly in \(\varepsilon\); values \(\varepsilon=10^{-5}\dots 10^{-6}\) in Hartree units are stated to be sufficient for chemical accuracy, and extrapolation \(\varepsilon\to0\) may be carried out by fitting energies to a low-order power law in \(\varepsilon\) [1606.08475].

The deterministic PCI study reports benchmark calculations on \(\mathrm{N}_2\) at equilibrium and stretched geometries, where chemical accuracy is achieved with wave functions containing less than \(0.5\%\) of the full CI space. It also reports computations on the ground state of \(\mathrm{C}_2\) with up to quaduple-\(\zeta\) basis sets and wave functions as large as 200 million determinants, allowing direct comparison with FCIQMC and density matrix renormalization group (DMRG). The wall-Chebyshev generator is noted to require only two vectors of storage and no three-vector recurrence, which the authors identify as advantageous for large-scale deterministic CI implementations [1606.08475].

## 4. Stochastic projector-based quantum Monte Carlo

A stochastic implementation of an approximate wall projector has been introduced for determinant-based QMC, specifically in initiator FCIQMC and multi-reference coupled-cluster Monte Carlo (MR-CCMC). In this setting, the projector is realized through the same truncated Chebyshev expansion,
\[
g_m(\hat H)=\sum_{n=0}^m c_n\,T_n(\tilde H),
\]
with coefficients chosen so that \(g_m(x)\approx \mathrm{wall}(x-x_0)\) over \([-1,1]\). Applying the three-term recurrence in operator form,
\[
|\Psi^{(n+1)}\rangle = 2\,\tilde H\,|\Psi^{(n)}\rangle - |\Psi^{(n-1)}\rangle,
\]
one obtains the projected state
\[
\sum_{n=0}^m c_n\,|\Psi^{(n)}\rangle.
\]
Each recurrence requires a single application of \(\hat H\), so an \(m\)th-order projection costs \(m\) Hamiltonian applications [2402.16685].

The stochastic algorithm is also expressed as a product of linear factors,
\[
\prod_{\nu=0}^{m-1}\Bigl[1-\tfrac{1}{a_\nu-S}(\hat H-S)\Bigr],
\]
with Chebyshev nodes
\[
a_\nu=E_0 + \tfrac{R}{2}\Bigl(1-\cos\!\bigl(\nu\pi/(m+\tfrac12)\bigr)\Bigr),
\]
and effective time steps \(\delta\tau_\nu=1/(a_\nu-S)\). In MR-CCMC, newly spawned determinants are screened by a metric-tree (BK-tree) search to determine whether they lie within the prescribed excitation level of the reference space, and symmetry screening may be applied at reference selection [2402.16685].

The convergence analysis states that the asymptotic convergence factor is
\[
\mu\approx 1-(E_1-E_0)\gamma,
\]
with \(\gamma=-g'(E_0)\) growing linearly with \(m\). Relative to the largest stable linear propagation step \(\delta\tau_{\max}\), the \(m\)th-order wall-Chebyshev projector is reported to provide a theoretical speed-up of \((m+1)/3\) in the number of projector steps. The same study states that, in practice, the Chebyshev weights avoid blooms, and observed speed-ups are larger; for \(m=5\), the wall-Chebyshev projector often reduces the required Hamiltonian applications by an order of magnitude or more to reach a given statistical error [2402.16685].

The reported benchmarks are correspondingly sharp. For \(\mathrm{Be}_2/\mathrm{cc}\text{-}\mathrm{pVQZ}\) with a \((4e,8o)\) CAS and MR-CCMCSD, a linear projector with \(\delta\tau\approx 0.002\) required \(\sim 3{,}034\) Hamiltonian applications to equilibrate to \(N_w\approx 3\times 10^6\) walkers, whereas the second-order Chebyshev projector (\(m=2\)) reached the same walker population with only 66 Hamiltonian applications. For \(\mathrm{C}_2/\mathrm{cc}\text{-}\mathrm{pVDZ}\), the first-order linear projector required \(\sim 500\) iterations to stabilize the shift and energy, while the fifth-order wall-Chebyshev projector (\(m=5\)) converged in \(\sim 50\) iterations and reduced wall-clock time from \(\sim 24\) h on 12 cores to \(\sim 2\) h on 6 cores for the same statistical error bar. The non-parallelity errors in the \(\mathrm{C}_2\) binding curve, relative to DMRG, are reported to drop by a factor of \(\gtrsim 2\) under wall-Chebyshev propagation [2402.16685].

## 5. Quantum algorithm for ground-state preparation

A quantum-algorithmic realization uses the same wall-Chebyshev series as a ground-state projector that is efficiently implemented as a product of Hamiltonian operators. In this formulation,
\[
P_K(H)=g_K^{\rm wall\!-\!Ch}(H)
=\frac{1}{1+2K}\sum_{k=0}^K(2-\delta_{k0})(-1)^k\,T_k(\tilde H),
\]
with
\[
\tilde H \equiv 2(H-E_0)/R -1,\qquad R\equiv E_{\max}-E_0.
\]
When \(E_0\) is not known exactly, one instead starts from an estimate \(S\ge E_0\), obtains an upper spectral estimate \(\widetilde E_{\max}\) from a Gershgorin circle bound,
\[
\widetilde E_{\max}=H_{jj}+\sum_{i\neq j}|H_{ij}|,
\]
chooses
\[
R=\alpha\,\widetilde E_{\max}-S,\qquad \alpha\ge 1,
\]
and works with
\[
\tilde H \equiv 1-2(H-S)/R.
\]
This guarantees that the true spectrum lies in \([S,\alpha\,\widetilde E_{\max}]\) [2508.00533].

A key structural identity is the factorization
\[
g_K^{\rm wall\!-\!Ch}(H)
=
\prod_{\nu=1}^K \frac{H-a_\nu}{E_0-a_\nu},
\]
with nodes
\[
a_\nu = E_0 + \tfrac{R}{2}\Bigl[1-\cos\!\frac{\nu\pi}{K+\tfrac12}\Bigr].
\]
This leads directly to an implementation by linear combinations of unitaries (LCU): each factor is block-encoded, ancillas are prepared and uncomputed around a SELECT operation for the Pauli terms of \(H\), and post-selection realizes \((H-a_\nu I)/\alpha_\nu\). The total number of Hamiltonian-oracle calls is \(O(K)\), and if \(H=\sum_{j=1}^L h_j P_j\), the total gate count is stated as \(\sim O(K\cdot L\cdot \mathrm{poly}(\log L))\) with \(O(\log L)\) ancilla overhead [2508.00533].

The error analysis is expressed in terms of the convergence factor
\[
\gamma
=
-\left.\frac{d\,g^{\rm wall\!-\!Ch}_K(E)}{dE}\right|_{E=E_0}
=
\frac{2K(K+1)}{3R},
\]
which yields suppression of the first excited component according to
\[
\left|\frac{g(E_1)}{g(E_0)}\right| \approx e^{-\Delta\gamma},
\qquad \Delta=E_1-E_0.
\]
The resulting asymptotic scaling is
\[
Q=K=O\!\bigl(\Delta^{-1/2}[\log(c_0^{-1}\epsilon^{-1})]^{1/2}\bigr),
\]
for query complexity, and
\[
G=O\!\bigl(K\cdot L\cdot \mathrm{poly}(\log L)\bigr)
\]
for gate complexity. With amplitude amplification, the query count acquires an additional factor \(O(P_{\mathrm{succ}}^{-1/2})\), where the success probability is
\[
P_{\mathrm{succ}}\simeq
\langle\psi|[g_m(H)]^\dagger g_m(H)|\psi\rangle/\alpha^{2K}.
\]
The analysis further states that in practice \(P_{\mathrm{succ}}=\Omega(c_0^2)\) [2508.00533].

The reported numerical benchmarks emphasize robustness to inaccurate ground-state-energy estimates. For hydrogen chains in STO-3G, the polynomial order \(K\) needed for 1 mHartree accuracy is tabulated for known and unknown \(E_0\), and the wall-Chebyshev projector maintains moderate orders in both regimes. For the two-site Hubbard model at \(U=1,5,10\), when \(E_0\) is known the wall-Chebyshev and eigenstate-filter projectors converge in \(K\approx 10\)–20 while step-function and ITE projectors need \(\approx 50\)–100; if \(E_0\) is unknown and set to Hartree–Fock, only wall-Chebyshev and ITE still converge. The same study states that eigenstate-filter and step-function projectors both rely critically on knowing \(E_0\) to within \(O(\Delta)\), whereas wall-Chebyshev remains the most slowly damped mode even when the shift is inaccurate [2508.00533].

## 6. Relation to other Chebyshev spectral projectors

The wall-Chebyshev projector belongs to a broader class of Chebyshev spectral filters, but the target subspace and approximation objective vary substantially across applications. In FEAST-style singular-value computation, the relevant object is not a ground-state wall projector but a step-function projector onto an interval \([a,b]\). There, the step function
\[
h(x)=
\begin{cases}
1,&x\in(a,b),\\
\tfrac12,&x\in\{a,b\},\\
0,&x\in[-1,1]\setminus[a,b]
\end{cases}
\]
is approximated by a Chebyshev–Jackson expansion
\[
\psi_d(x)=\frac{c_0}{2}+\sum_{j=1}^d \rho_{j,d}\,c_j\,T_j(x),
\]
with Jackson damping factors \(\rho_{j,d}\) used to suppress Gibbs oscillations. The resulting approximate spectral projector is symmetric positive semi-definite with eigenvalues in \([0,1]\), admits pointwise error bounds of order \((d+2)^{-3}\), and is used in a FEAST SVDsolver without shifted linear systems; convergence is governed by the spectral gap of the approximate projector, with subspace convergence rate \((\gamma_{p+1}/\gamma_p)^k\) [2201.02901].

In large-scale Kohn–Sham density functional theory, the relevant object is the occupied-subspace projector rather than a ground-state-only projector. The two-level Chebyshev filter based complementary subspace method (CS2CF) uses an outer Chebyshev filter on the full Hamiltonian to compute a basis \(Y\) for the occupied subspace and an inner Chebyshev filter on the projected Hamiltonian \(\tilde H=Y^T H Y\) to compute only the \(N_t\) fractionally occupied states. The density projector in the projected space is written as
\[
\tilde P = I-\tilde C\tilde C^T,
\]
so that the full density matrix is
\[
P = Y\tilde P Y^T = YY^T-(Y\tilde C)(Y\tilde C)^T.
\]
Because this avoids full diagonalization of \(\tilde H\), the dominant \(O(N_s^3)\) bottleneck is reduced to
\[
O(\tilde m\,N_s^2N_t + N_sN_t^2 + N_t^3),
\qquad N_t\ll N_s.
\]
For an 8,000-atom bulk silicon system with 32,000 electrons on 34,560 cores, the reported timings are approximately 9 s for the outer ALB update, 2 s for the Hamiltonian-vector update, and 40 s for the CS2CF subspace solve, for a total SCF time of approximately 51 s; a full ELPA diagonalization of the \(32000\times 32000\) DG matrix is reported to take approximately 650 s per SCF on the same cores. The same implementation yields 1.0 ps of ab initio molecular dynamics in approximately 28 h of wall time, with an average SCF step wall time of 51 s [1712.04439].

These related methods clarify a common misconception. “Chebyshev projector” does not denote a single algorithmic object. In the wall-Chebyshev setting, the polynomial approximates the infinite-imaginary-time wall that isolates the ground state; in Chebyshev–Jackson FEAST, it approximates an interval indicator; in CS2CF, it constructs occupied-subspace and complementary-subspace projectors for electronic structure. The common element is the use of explicitly controlled polynomial filtering on a spectrally rescaled operator, but the spectral target, convergence criterion, and implementation pathway differ materially across these settings [2201.02901; 1712.04439].

Source: https://www.emergentmind.com/topics/wall-chebyshev-projector