Comment on "Spin-1/2 Kagome Heisenberg Antiferromagnet: Machine Learning Discovery of the Spinon Pair-Density-Wave Ground State"
For researchers studying quantum spin liquids and frustrated magnets, this comment corrects a potentially misleading result by identifying a methodological flaw in the original paper.
This comment re-analyzes a recent study claiming a spinon pair-density-wave ground state for the kagome Heisenberg antiferromagnet using neural networks. The authors show that the reported low energies are artifacts of broken ergodicity in Metropolis-Hastings sampling, and with proper ergodic sampling the neural network energies are higher than DMRG results, invalidating the original claims.
A recent article [Phys. Rev. X 15, 011047 (2025)] utilizes group-equivariant convolutional neural networks to study the ground state of the kagome Heisenberg antiferromagnet. On the largest finite-size cluster studied to date ($N=108$), the authors report variational energies significantly lower than other numerical methods, including state-of-the-art density matrix renormalization group (DMRG) calculations. In contrast to previous results suggesting a possible spin-liquid ground state, the authors observe a spinon pair-density-wave ground state. We find that: (i) the reported low energies are artifacts of broken ergodicity in the Metropolis--Hastings sampling, since the single-spin-flip update rule utilized by the authors effectively freezes the Markov chains; and (ii) when ergodic sampling is enforced via spin-exchange updates, the neural network converges to energies significantly higher than existing DMRG results, calling the paper's claims into question.