Abstract:We consider a fermionic system for which there exist a single-reference configuration-interaction (CI) expansion of the ground state wave function that converges, albeit not necessarily rapidly, with respect to excitation number. We show that, if the coefficients of Slater determinants (SD) with $l\leq k$ excitations can be defined with a number of free parameters $N_{\leq k}$ bounded polynomially in $k$, the ground state energy $E$ only depends on a small fraction of all the wave function parameters, and is the solution of equations of the coupled-cluster (CC) form. This generalizes the standard CC method, for which $N_{\leq k}$ is bounded by a constant. Based on that result and low-rank tensor decompositions (LRTD), we discuss two possible extensions of the CC approach for wave functions with general polynomial bound for $N_{\leq k}$. First, one can use LRTD to represent the amplitudes of the CC cluster operator $T$ which, unlike in the CC case, is not truncated with respect to excitation number, and the energy and tensor parameters are given by a LRTD-adapted version of standard CC equations. Second, the LRTD can also be used to directly parametrize the CI coefficients, which involves different equations of the CC form. We derive those equations for up to quadruple-excitation coefficients, using a different type of excitation operator in the CC wave function ansatz, and a Hamiltonian representation in terms of excited particle and hole operators. We complete the proposed CC extensions by constructing compact tensor representations of coefficients, or $T$-amplitudes, using superpositions of tree tensor networks which take into account different possible types of entanglement between excited particles and holes. Finally, we discuss why the proposed extensions are theoretically applicable at larger coupling strengths than those treatable by the standard CC method.

Robust Tensor Hypercontraction of the Particle-Particle Ladder Term in Equation-of-Motion Coupled Cluster Theory

Robust Tensor Hypercontraction of the Particle-Particle Ladder Term in Equation-of-Motion Coupled Cluster Theory

Tensor Hypercontraction of Cluster Perturbation Theory: Quartic Scaling Perturbation Series for the Coupled Cluster Singles and Doubles Ground-State Energies

Open-shell Tensor Hypercontraction

Enhanced Krylov Methods for Molecular Hamiltonians: Reduced Memory Cost and Complexity Scaling via Tensor Hypercontraction

Even Faster Exact Exchange for Solids via Tensor Hypercontraction

An effective, "strong" electron correlation estimate using alternative formulations of T 2 -based coupled cluster theory and the factorization theorem

Compression of the electron repulsion integral tensor in tensor hypercontraction format with cubic scaling cost

Spatial Signatures of Electron Correlation in Least-Squares Tensor Hyper-Contraction

Superpositions of tree-tensor networks for single-reference ground states in the strong correlation regime

An "ultimate" coupled cluster method based entirely on $T_2$

Tensor Product Approximation (DMRG) and Coupled Cluster method in Quantum Chemistry

Hyperoptimized Approximate Contraction of Tensor Networks with Arbitrary Geometry

Automatic transformation of irreducible representations for efficient contraction of tensors with cyclic group symmetry

Computations in quantum tensor networks

Improved Grid Optimization and Fitting in Least Squares Tensor Hypercontraction

Energy of fermionic ground states with low-entanglement single-reference expansions, and tensor-based strong-coupling extensions of the coupled-cluster method

Hyper-optimized approximate contraction of tensor networks with arbitrary geometry

A framework for load balancing of tensor contraction expressions via dynamic task partitioning

Low-Scaling Algorithm for the Random Phase Approximation using Tensor Hypercontraction with k-point Sampling

Reducing Computational Complexity of Tensor Contractions via Tensor-Train Networks