Flow-field inference from neural data using deep recurrent networks.

Computations involved in processes such as decision-making, working memory, and motor control are thought to emerge from the dynamics governing the collective activity of neurons in large populations. But the estimation of these dynamics remains a significant challenge. Here we introduce Flow-field Inference from Neural Data using deep Recurrent networks (FINDR), an unsupervised deep learning method that can infer low-dimensional nonlinear stochastic dynamics underlying neural population activity. Using population spike train data from frontal brain regions of rats performing an auditory decision-making task, we demonstrate that FINDR outperforms existing methods in capturing the heterogeneous responses of individual neurons. We further show that FINDR can discover interpretable low-dimensional dynamics when it is trained to disentangle task-relevant and irrelevant components of the neural population activity. Importantly, the low-dimensional nature of the learned dynamics allows for explicit visualization of flow fields and attractor structures. We suggest FINDR as a powerful method for revealing the low-dimensional task-relevant dynamics of neural populations and their associated computations.

Theory of Gating in Recurrent Neural Networks.

Recurrent neural networks (RNNs) are powerful dynamical models, widely used in machine learning (ML) and neuroscience. Prior theoretical work has focused on RNNs with additive interactions. However gating i.e., multiplicative interactions are ubiquitous in real neurons and also the central feature of the best-performing RNNs in ML. Here, we show that gating offers flexible control of two salient features of the collective dynamics: (i) timescales and (ii) dimensionality. The gate controlling timescales leads to a novel marginally stable state, where the network functions as a flexible integrator. Unlike previous approaches, gating permits this important function without parameter fine-tuning or special symmetries. Gates also provide a flexible, context-dependent mechanism to reset the memory trace, thus complementing the memory function. The gate modulating the dimensionality can induce a novel, discontinuous chaotic transition, where inputs push a stable system to strong chaotic activity, in contrast to the typically stabilizing effect of inputs. At this transition, unlike additive RNNs, the proliferation of critical points (topological complexity) is decoupled from the appearance of chaotic dynamics (dynamical complexity). The rich dynamics are summarized in phase diagrams, thus providing a map for principled parameter initialization choices to ML practitioners.

Spectral Gaps and Midgap States in Random Quantum Master Equations.

We discuss the decay rates of chaotic quantum systems coupled to noise. We model both the Hamiltonian and the system-noise coupling by random N×N Hermitian matrices, and study the spectral properties of the resulting Liouvillian superoperator. We consider various random-matrix ensembles, and find that for all of them the asymptotic decay rate remains nonzero in the thermodynamic limit; i.e., the spectrum of the superoperator is gapped as N→∞. For finite N, the probability of finding a very small gap vanishes as P(Δ)∼Δ^{cN}, where c is insensitive to the dissipation strength. A sharp spectral transition takes place as the dissipation strength is increased: for dissipation beyond a critical strength, the slowest-decaying eigenvalues of the Liouvillian correspond to isolated "midgap" states. We give evidence that midgap states exist also for nonrandom system-noise coupling and discuss some experimental implications of the above results.

Electromagnetic and gravitational responses of photonic Landau levels.

Topology has recently become a focus in condensed matter physics, arising in the context of the quantum Hall effect and topological insulators. In both of these cases, the topology of the system is defined through bulk properties ('topological invariants') but detected through surface properties. Here we measure three topological invariants of a quantum Hall material-photonic Landau levels in curved space-through local electromagnetic and gravitational responses of the bulk material. Viewing the material as a many-port circulator, the Chern number (a topological invariant) manifests as spatial winding of the phase of the circulator. The accumulation of particles near points of high spatial curvature and the moment of inertia of the resultant particle density distribution quantify two additional topological invariants-the mean orbital spin and the chiral central charge. We find that these invariants converge to their global values when probed over increasing length scales (several magnetic lengths), consistent with the intuition that the bulk and edges of a system are distinguishable only for sufficiently large samples (larger than roughly one magnetic length). Our experiments are enabled by applying quantum optics tools to synthetic topological matter (here twisted optical resonators). Combined with advances in Rydberg-mediated photon collisions, our work will enable precision characterization of topological matter in photon fluids.

Publisher's Note: Particle-Hole Duality in the Lowest Landau Level [Phys. Rev. Lett. 118, 206602 (2017)].

This corrects the article DOI: 10.1103/PhysRevLett.118.206602.

Publisher's Note: Particle-Hole Duality in the Lowest Landau Level [Phys. Rev. Lett. 118, 206602 (2017)].

This corrects the article DOI: 10.1103/PhysRevLett.118.206602.

Particle-Hole Duality in the Lowest Landau Level.

We derive a number of exact relations between response functions of holomorphic, chiral fractional quantum Hall states and their particle-hole (PH) conjugates. These exact relations allow one to calculate the Hall conductivity, Hall viscosity, various Berry phases, and the static structure factor of PH conjugate states from the corresponding properties of the original states. These relations establish a precise duality between chiral quantum Hall states and their PH conjugates. The key ingredient in the proof of the relations is a generalization of Girvin's construction of PH-conjugate states to inhomogeneous magnetic field and curvature. Finally, we make several nontrivial checks of the relations, including for the Jain states and their PH conjugates.

Coherent transport and energy flow patterns in photosynthesis under incoherent excitation.

Long-lived coherences have been observed in photosynthetic complexes after laser excitation, inspiring new theories regarding the extreme quantum efficiency of photosynthetic energy transfer. Whether coherent (ballistic) transport occurs in nature and whether it improves photosynthetic efficiency remain topics of debate. Here, we use a nonequilibrium Green's function analysis to model exciton transport after excitation from an incoherent source (as opposed to coherent laser excitation). We find that even with an incoherent source, the rate of environmental dephasing strongly affects exciton transport efficiency, suggesting that the relationship between dephasing and efficiency is not an artifact of coherent excitation. The Green's function analysis provides a clear view of both the pattern of excitonic fluxes among chromophores and the multidirectionality of energy transfer that is a feature of coherent transport. We see that even in the presence of an incoherent source, transport occurs by qualitatively different mechanisms as dephasing increases. Our approach can be generalized to complex synthetic systems and may provide a new tool for optimizing synthetic light harvesting materials.

Atomic resolution imaging of currents in nanoscopic quantum networks via scanning tunneling microscopy.

We propose a new method for atomic-scale imaging of spatial current patterns in nanoscopic quantum networks by using scanning tunneling microscopy (STM). By measuring the current flowing from the STM tip into one of the leads attached to the network as a function of tip position, one obtains an atomically resolved spatial image of "current riverbeds" whose spatial structure reflects the coherent flow of electrons out of equilibrium. We show that this method can be successfully applied in a variety of network topologies and is robust against dephasing effects.