arXiv Science⌕ Search
Home/Archive/arXiv

arXiv · 2610.35934

Agentic Swarms Discover Enhanchment of Fractional Chern Insulator Gaps Beyond Ideal Quantum Geometry

Yitong Zhang·Changwen Xu·Andrew Boshi Li·Siddhartha Sarkar·Kai Sun·Venkatasubramanian Viswanathan
View original source ↗Read source document ↗Export metadata ↓

Abstract

Can artificial intelligence (AI) move beyond solving well-defined problems to contribute to open-ended discovery in fundamental physics? Could AI propose what to explore, identify unexpected phenomena, and uncover their mechanisms? We address this using MOSAIC (Multi-agent Open-ended Scientific Analysis and Investigation through Collaboration), a collaborative swarm of AI agents (Claude Opus 5.5), to tackle a longstanding bottleneck: the small energy gaps limiting the robustness of fractional topological order and the study of anyons at higher temperatures. Rather than prescribing a model, we ask frontier models (GPT-6 Astra) to propose a theoretical framework, which MOSAIC explores through diverse strategies and rapid exchange of ideas. This reveals a striking finding: although ideal quantum geometry is considered favorable for fractional states, departing from this limit can dramatically enhance their robustness, with many-body gaps more than an order of magnitude larger than in the ideal case. Without centralized coordination, the agents explored autonomously, combined findings across agents, and formed insights from individual runs, outperforming isolated agents with the same tools. AI-assisted theoretical analysis further develops simplified models with an analytical gap enhancement factor that is unbounded within the single-band approximation. Remarkably, AI identifies the real-space quantum metric, rather than the momentum-space one, as the key quantity governing the many-body gap, revealing how flat-band wavefunctions shape two-electron interactions and yielding an index computable from single-particle wavefunctions without many-body ED to guide gap optimization. Our study demonstrates how collaborative AI agents can integrate exploration, discovery of unexpected physics, and theoretical understanding, opening new possibilities for AI-driven research in fundamental physics.

Explore related subjects

cond-mat.str-el

Keep this discovery

Explore connections, maps & timelines

BibTeXRIS

Yitong Zhang, Changwen Xu, Andrew Boshi Li, Siddhartha Sarkar, Kai Sun, Venkatasubramanian Viswanathan. 2026-10-20. Agentic Swarms Discover Enhanchment of Fractional Chern Insulator Gaps Beyond Ideal Quantum Geometry. https://arxiv.org/abs/2610.35934

Cite the original work for its findings. Save a collection to share your selection of sources.

KEEP EXPLORING

Related papers

arXiv2026

Point-Group Equivariant Neural Networks for Local Dynamics in Lattice Systems

We develop point-group equivariant neural networks for learning local mappings used in dynamical simulations of lattice systems. The same network acts on a finite neighborhood around each site, respecting lattice translations through shared parameters. The local configuration is decomposed into features that transform according to irreducible representations of the point group, and hidden features retain this organization throughout the network. Equivariant linear maps and symmetry-allowed tensor products then learn nonlinear spatial correlations while preserving the required transformation laws. This approach avoids constructing a fixed set of hand-crafted invariant descriptors, whose selection requires choices about which local correlations to retain. It also allows the complexity of the learned mapping to grow without enlarging its spatial cutoff. We demonstrate the construction by learning the forces governing adiabatic dynamics in the square-lattice Holstein model. Comparisons with exact diagonalization test the accuracy of the predicted forces and the evolution of spatial correlations. Force evaluation scales linearly with system size, allowing the trained model to simulate charge-density-wave ordering on large lattices. These simulations reproduce anomalously slow coarsening of the charge-density-wave domains, illustrating the use of the learned local mapping for collective dynamics.

cond-mat.str-el↗
arXiv2026

Topological Chiral Superconductivity from Doping Correlated Charge Transfer Insulator with Orbital Magnetism

Topological chiral superconductivity has been extensively studied from the perspective of electronic structures in momentum space, while its emergence from real-space processes in strongly interacting systems remains largely unexplored. Here, we study a doped charge-transfer insulator in a strongly repulsive spinless-fermion model with three sublattices and $2π$ flux per unit cell. Combining perturbative analysis, exact diagonalization, and density matrix renormalization group simulations, we establish a topological superconducting ground state with $p+ip$ symmetry. In the large charge-transfer gap limit, chiral charge-transfer excitations generate a correlated-hopping process that provides a microscopic mechanism for electron pairing and its chiral structure. The chiral superconductivity persists into the smaller-gap regime, where chiral excitonic two-particle bound state can emerge. Our results establish a concrete real-space route to topological chiral superconductivity upon doping strongly interacting charge-transfer insulators with orbital magnetism.

cond-mat.str-el↗
arXiv2026

Solving fractional quantum Hall problems with thermodynamic-limit bootstrap

Determining correlated ground states directly in the thermodynamic limit remains a central challenge in quantum many-body physics. Here we develop a quantum many-body bootstrap that yields a rigorous and tight lower bound on the thermodynamic ground-state energy of a fractional quantum Hall system together with an optimized reduced density matrix $ρ_W^\star$ on a spatial window. For sufficiently tight bounds, we find that $ρ_W^\star$ accurately reproduces the ground-state reduced density matrix, as directly tested by its fidelity with known model states. For exactly solvable Laughlin and Moore--Read parent Hamiltonians, the bootstrap also reproduces pair correlations and orbital entanglement spectra (OES). For Landau-level-projected Coulomb interactions across multiple fillings, the energy bounds closely track established thermodynamic estimates, while the window density matrix recovers the familiar $ν=1/3$ Laughlin and $ν=5/2$ Moore--Read states and the composite Fermi liquid at $ν=1/2$. More importantly, the same framework also resolves unsettled regimes. At $ν=1/4$, the OES and real-space correlations identify a composite Fermi liquid, while at $ν=12/5$ the combined diagnostics suggest $\overline{\mathrm{RR}}_3$ as the ground-state order. Our bootstrap therefore provides direct access to both ground-state energies and many-body structure in the thermodynamic limit.

cond-mat.str-el↗