Beyond Perturbation Theory: A Resolvent-Based Framework for Strongly Correlated Many-Body Systems

Abstract: Traditional perturbation theory, based on local analyticity (Taylor expansion), often fails in many-body systems with exponentially small energy gaps and strong interactions. This work presents an alternative methodological framework built on two core principles: (1) starting from the pole expansion of the resolvent to directly capture the global analytic structure, and (2) treating local fluctuations statistically (in the spirit of the eigenstate thermalization hypothesis) to close the mean-field equations. Crucially, we go beyond the mean-field level by deriving an exact recursive re-expansion of the cross-correlated terms, which systematically generates higher-order corrections that control the distribution tails, branch splitting, and fluctuations. The framework is realized through a hierarchical ansatz strategy, solving self-consistent equations with Lorentzian, Gaussian, and hybrid forms to describe the bulk, tail, and full distribution, respectively. This methodology does not rely on weak-coupling assumptions and is applicable to the quantitative analysis of global properties such as entropy production and distribution functions in nonintegrable many-body systems. We detail its mathematical structure, the recursive expansion of fluctuations, conditions of validity, comparison with traditional methods, and provide a general implementation workflow.

• The paper introduces a resolvent framework that bypasses divergence issues inherent to perturbative methods in strongly correlated systems.
• It employs a recursive expansion and hierarchical ansatz to capture global spectral properties including tails and branch splitting.
• The methodology enables quantitative predictions for observables like entropy and eigenstate statistics beyond conventional diagrammatic approaches.

Resolvent-Based Framework for Strongly Correlated Many-Body Systems: Technical Analysis

Motivation and Limitations of Conventional Perturbation Theory

The paper "Beyond Perturbation Theory: A Resolvent-Based Framework for Strongly Correlated Many-Body Systems" (2604.00606) addresses fundamental limitations of traditional perturbative techniques in quantum many-body theory. In strongly correlated settings, especially nonintegrable systems with exponentially dense spectra, Taylor-expansion based perturbation theory fails due to divergence, inability to capture global spectral features, and breakdown in the presence of strong interactions. Many physical observables, including entropy production and eigenstate statistics, depend on the global analytic properties of the Hamiltonian, not merely local behaviors.

Diagrammatic expansions (e.g., Dyson series, SCBA), although widely used, rely on truncations and typically describe sequential scattering along single propagation lines, ignoring multi-path effects. Random Matrix Theory (RMT), while modeling fully chaotic spectra, neglects structured randomness originating from the system's actual Hamiltonian. The Eigenstate Thermalization Hypothesis (ETH) provides a statistical mechanism for local fluctuations but lacks a predictive, quantitative framework for global distributions.

Core Methodological Principles

The authors introduce a resolvent-based formalism founded on four principles:

1. Global Analyticity via Pole Expansion: The resolvent $(z-H)^{-1}$ encodes spectral information through its poles (eigenvalues). The derived self-consistent equations capture global analytic structure, enabling direct formulation of spectral distribution functions without perturbative truncation.
2. Statistical Treatment of Fluctuations: Taking inspiration from ETH, off-diagonal matrix elements are modeled as random variables; cross-correlated terms average out by phase cancellation, allowing closed mean-field equations.
3. Recursive Expansion of Fluctuations: Exact re-expansion of cross-correlated terms (multi-resolvent products) provides systematic corrections beyond mean field. This hierarchy controls tails, branch splitting, and higher-order moments of spectral distributions.
4. Hierarchical Ansatz Strategy: Self-consistent equations are solved by physically motivated ansatzes: Lorentzian for the bulk, Gaussian for tails, and Voigt (hybrid Lorentzian-Gaussian) for full distributions. This enables accurate quantitative characterization across regimes.

Mathematical Structuring and Self-Consistency

The formalism is rooted in the resolvent operator:

$$\mathcal{R}_{\mu i}(z) = \langle \phi_{\mu i} | (z-H)^{-1} | \phi_{\mu i} \rangle$$

which is analytic except at eigenvalues $\lambda_n$. The central self-consistent equation, derived using projection identities, takes the form:

$$\mathcal{R}_{\mu i}(z) = \frac{1}{z - a_{\mu i} - V_{\mu i} - \mathcal{G}_{\mu i}(z)}$$

where $a_{\mu i}$ and $V_{\mu i}$ are eigenvalues and diagonal interactions, and $\mathcal{G}_{\mu i}(z)$ is the self-energy. The mean-field approximation applies statistical averaging to off-diagonal elements, yielding:

$$\mathcal{G}_{\mu i}(z) = \sum_{\nu j \neq \mu i} |V_{\mu i,\nu j}|^2 \mathcal{R}_{\nu j}(z)$$

and connects the imaginary part of $\mathcal{R}_{\mu i}(z)$ to the spectral probability distribution.

Beyond mean field, the recursive expansion introduces multi-resolvent terms (e.g., third-order terms $\mathcal{G}^{(3)}$). The paper rigorously derives how these corrections generate non-trivial, nonlocal Hilbert transform contributions to the spectral distributions, resulting in asymmetry, Gaussian tails, and branch splitting, as explicitly shown for nonintegrable Ising models.

Hierarchical Ansatz and Effective Self-Energy Representations

The ansatz strategy for the spectral distributions $$\mathcal{R}_{\mu i}(z) = \langle \phi_{\mu i} | (z-H)^{-1} | \phi_{\mu i} \rangle$$0 is realized as follows:

• Lorentzian Ansatz: Appropriate in regions of large spectral weight; normalization and self-consistent parameter equations derive from matching the global behavior of the resolvent.
• Gaussian Ansatz: Accurate for distribution tails when energy differences are large, with self-consistency established through statistical averaging and exponential decay of interaction matrix elements.
• Voigt (Lorentzian-Gaussian) Ansatz: Unified description incorporating both bulk and tail features, with normalization ensured by integral constraints.

The effective self-energy formalism generalizes these ansatzes, introducing frequency-dependent imaginary parts via Faddeeva functions. This construction inherently fulfills Kramers–Kronig relations, ensuring causality and normalization, and is analytically tractable for self-consistent determination.

Structural Distinction from Diagrammatic Methods

The resolvent hierarchy fundamentally differs from finite diagrammatic resummations (e.g., SCBA, ladder, parquet) in its algebraic generation rules. While diagrammatic self-energies are additive and limited to sequential, single-chain processes, the resolvent hierarchy encapsulates products of multiple full resolvents. Recursive substitution generates nested hierarchies unbounded by any finite skeleton topology, leading to convolution-type nonlocal frequency dependencies. This structure is essential for capturing statistical fluctuations, tails, and branch-splitting phenomena in strongly correlated systems.

Applicability, Generalization, and Limitations

The framework is valid under generic nonintegrability, random-phase conditions, and exponentially large Hilbert spaces, typical of quantum chaotic systems. It can be adapted to fermionic/bosonic operators, out-of-equilibrium dynamics, open quantum systems, and cases with additional symmetries.

Limitations arise when statistical assumptions fail, e.g., near integrability or in systems with strong symmetry constraints. The methodology requires coarse-graining and self-consistent closure, which may be computationally intensive for large hierarchies.

Qualitative and Quantitative Predictions

Strong numerical results confirm the framework's accuracy for bulk and tail distributions, spectral asymmetry, and Gaussian tail behavior. Contradictory to conventional wisdom, even in strongly interacting regimes, Lorentzian spectral functions can be systematically corrected by recursive expansions, yielding unified analytic control over the full spectral line shapes.

Implications and Future Directions

Practically, the framework enables quantitative predictions for global properties such as entropy and distribution functions in nonintegrable many-body systems previously inaccessible to perturbative or diagrammatic approaches. Theoretically, it reorganizes many-body physics at the level of pole structure and analytic properties, directly connecting microscopic interactions to emergent statistical phenomena.

Future developments may include adaptive ansatz schemes for more complex systems, analytic characterization of higher-order corrections, and generalizations to non-equilibrium or open quantum settings. The explicit connection to Kramers–Kronig relations and analytic continuation opens pathways for direct calculation of time-dependent Green's functions and quantum transport observables.

Conclusion

The resolvent-based framework presented provides a unified, analytic, and systematic methodology for the study of strongly correlated many-body systems, transcending the limitations of traditional perturbation theory and diagrammatic resummations. The recursive hierarchy of multi-resolvent terms establishes direct links between microscopic cross-correlations and global spectral features, offering controlled quantitative predictions and robust theoretical insight into quantum thermalization, entropy production, and statistical mechanics of complex quantum systems (2604.00606).