An optimized construction of lie algebra generator pools for variational quantum eigensolvers in chemistry

The evolution of quantum states relies on unitary transformations dictated by Lie algebras. Finding a minimal set of generators to construct these algebras, known as a minimal complete pool (MCP), is a central challenge in the field. Traditionally, the search for these generators has relied on greedy construction steps applied to an exponentially growing number of candidate operators, creating an intractable computational bottleneck for even moderately sized systems.

A robust new mathematical framework addresses these challenges head-on, delivering a polynomial-scaling strategy that bypasses the old computational walls. By translating the algebraic properties of the generators into a clean vector space representation, this approach is more robust and reliable, unlocking highly efficient variational quantum eigensolver (VQE) simulations for real chemistry workflows.

• The Leap in VQE Performance, Powered by Linear Algebra.
  The standout gain is computational efficiency. Rather than brute-forcing the explicit construction of the dynamical Lie algebra, the new approach maps Pauli operators to a binary vector space. By translating operator anti-commutation relations into a symmetric binary matrix, the question of whether a pool is mathematically complete boils down to evaluating the rank and congruence of this single matrix. Instead of relying on naive exponential calculations or unverified polynomial heuristics, this elegant method reduces the verification workload to a mathematically provable O(N³) scaling, where N is the number of qubits.
•     Streamlined Workflow and Reduced Measurement Overhead.
  The new method improves the workflow of adaptive variational quantum algorithms. Standard iterative approaches like ADAPT-VQE require evaluating the energy gradients of every candidate operator, which can impose a punishing O(N⁸) measurement scaling when using traditional O(N⁴) pools. By deploying mathematically guaranteed, linearly scaling MCPs, the new MB-ADAPT-VQE framework, using these MCPs, overcomes this measurement overhead. This allows teams to iterate more quickly, simulating complex, strongly correlated systems, like a 26-qubit H₂O molecule, with far fewer resources. It also removes the initial computational hurdles that previously restricted fixed-ansatz methods like NI-DUCC-VQE to small molecules using MCPs, allowing them to be applied to much larger targets.
•     Greater Expressiveness with Physics-Informed Starters.
  While the new theoretical framework is powerful, mathematical completeness alone is not the whole story. If an operator pool ignores inherent molecular symmetries, the algorithm will crash into a "symmetry roadblock" with vanishing gradients. To solve this, the mathematically minimal core is repopulated with a robust set of physically-motivated "starters". This hybrid strategy expands the scope for tackling complex chemical systems: it guarantees Lie-algebraic completeness while providing the classical optimizer with the expressive physical pathways needed to efficiently reach the true ground state.

In summary, this approach shifts pool construction from a costly trial-and-error process to a highly efficient framework. Showing that smaller physics-informed complete operator pools can speed up simulations provides a solid foundation for quantum chemistry algorithms, with broader utility in areas like quantum error correction and machine learning.