Search

Virtual, make-on-demand chemical libraries have transformed early-stage drug discovery by unlocking vast, synthetically accessible regions of chemical space. Recent years have witnessed rapid growth in these libraries from millions to trillions of compounds, hiding undiscovered, potent hits for a variety of therapeutic targets. However, they are quickly approaching a size beyond that which permits explicit enumeration, presenting new challenges for virtual screening. To overcome these challenges, we propose the Combinatorial Synthesis Library Variational Auto-Encoder (CSLVAE). The proposed generative model represents such libraries as a differentiable, hierarchically-organized database. Given a compound from the library, the molecular encoder constructs a query for retrieval, which is utilized by the molecular decoder to reconstruct the compound by first decoding its chemical reaction and subsequently decoding its reactants. Our design minimizes autoregression in the decoder, facilitating the generation of large, valid molecular graphs. Our method performs fast and parallel batch inference for ultra-large synthesis libraries, enabling a number of important applications in early-stage drug discovery. Compounds proposed by our method are guaranteed to be in the library, and thus synthetically and cost-effectively accessible. Importantly, CSLVAE can encode out-of-library compounds and search for in-library analogues. In experiments, we demonstrate the capabilities of the proposed method in the navigation of massive combinatorial synthesis libraries., Comment: 36th Conference on Neural Information Processing Systems (NeurIPS 2022)

We introduce a new setting, optimize-and-estimate structured bandits. Here, a policy must select a batch of arms, each characterized by its own context, that would allow it to both maximize reward and maintain an accurate (ideally unbiased) population estimate of the reward. This setting is inherent to many public and private sector applications and often requires handling delayed feedback, small data, and distribution shifts. We demonstrate its importance on real data from the United States Internal Revenue Service (IRS). The IRS performs yearly audits of the tax base. Two of its most important objectives are to identify suspected misreporting and to estimate the "tax gap" -- the global difference between the amount paid and true amount owed. Based on a unique collaboration with the IRS, we cast these two processes as a unified optimize-and-estimate structured bandit. We analyze optimize-and-estimate approaches to the IRS problem and propose a novel mechanism for unbiased population estimation that achieves rewards comparable to baseline approaches. This approach has the potential to improve audit efficacy, while maintaining policy-relevant estimates of the tax gap. This has important social consequences given that the current tax gap is estimated at nearly half a trillion dollars. We suggest that this problem setting is fertile ground for further research and we highlight its interesting challenges. The results of this and related research are currently being incorporated into the continual improvement of the IRS audit selection methods., Comment: Accepted to the Thirty-Seventh AAAI Conference On Artificial Intelligence (AAAI), 2023

Concentrated Animal Feeding Operations (CAFOs) pose serious risks to air, water, and public health, but have proven to be challenging to regulate. The U.S. Government Accountability Office notes that a basic challenge is the lack of comprehensive location information on CAFOs. We use the USDA's National Agricultural Imagery Program (NAIP) 1m/pixel aerial imagery to detect poultry CAFOs across the continental United States. We train convolutional neural network (CNN) models to identify individual poultry barns and apply the best performing model to over 42 TB of imagery to create the first national, open-source dataset of poultry CAFOs. We validate the model predictions against held-out validation set on poultry CAFO facility locations from 10 hand-labeled counties in California and demonstrate that this approach has significant potential to fill gaps in environmental monitoring.

We explore the promises and challenges of employing sequential decision-making algorithms -- such as bandits, reinforcement learning, and active learning -- in law and public policy. While such algorithms have well-characterized performance in the private sector (e.g., online advertising), the tendency to naively apply algorithms motivated by one domain, often online advertisements, can be called the "advertisement fallacy." Our main thesis is that law and public policy pose distinct methodological challenges that the machine learning community has not yet addressed. Machine learning will need to address these methodological problems to move "beyond ads." Public law, for instance, can pose multiple objectives, necessitate batched and delayed feedback, and require systems to learn rational, causal decision-making policies, each of which presents novel questions at the research frontier. We discuss a wide range of potential applications of sequential decision-making algorithms in regulation and governance, including public health, environmental protection, tax administration, occupational safety, and benefits adjudication. We use these examples to highlight research needed to render sequential decision making policy-compliant, adaptable, and effective in the public sector. We also note the potential risks of such deployments and describe how sequential decision systems can also facilitate the discovery of harms. We hope our work inspires more investigation of sequential decision making in law and public policy, which provide unique challenges for machine learning researchers with potential for significant social benefit., Comment: Version 1 presented at Causal Inference Challenges in Sequential Decision Making: Bridging Theory and Practice (2021), a NeurIPS 2021 Workshop; Version 2 presented at the 2nd ACM Symposium on Computer Science and Law (2022) (DOI: https://dl.acm.org/doi/10.1145/3511265.3550439)

In drug discovery, structure-based virtual high-throughput screening (vHTS) campaigns aim to identify bioactive ligands or "hits" for therapeutic protein targets from docked poses at specific binding sites. However, while generally successful at this task, many deep learning methods are known to be insensitive to protein-ligand interactions, decreasing the reliability of hit detection and hindering discovery at novel binding sites. Here, we overcome this limitation by introducing a class of models with two key features: 1) we condition bioactivity on pose quality score, and 2) we present poor poses of true binders to the model as negative examples. The conditioning forces the model to learn details of physical interactions. We evaluate these models on a new benchmark designed to detect pose-sensitivity., Comment: Machine Learning for Structural Biology, NeurIPS 2021

Much environmental enforcement in the United States has historically relied on either self-reported data or physical, resource-intensive, infrequent inspections. Advances in remote sensing and computer vision, however, have the potential to augment compliance monitoring by detecting early warning signs of noncompliance. We demonstrate a process for rapid identification of significant structural expansion using Planet's 3m/pixel satellite imagery products and focusing on Concentrated Animal Feeding Operations (CAFOs) in the US as a test case. Unpermitted building expansion has been a particular challenge with CAFOs, which pose significant health and environmental risks. Using new hand-labeled dataset of 145,053 images of 1,513 CAFOs, we combine state-of-the-art building segmentation with a likelihood-based change-point detection model to provide a robust signal of building expansion (AUC = 0.86). A major advantage of this approach is that it can work with higher cadence (daily to weekly), but lower resolution (3m/pixel), satellite imagery than previously used in similar environmental settings. It is also highly generalizable and thus provides a near real-time monitoring tool to prioritize enforcement resources in other settings where unpermitted construction poses environmental risk, e.g. zoning, habitat modification, or wetland protection.

While self-supervised learning has made rapid advances in natural language processing, it remains unclear when researchers should engage in resource-intensive domain-specific pretraining (domain pretraining). The law, puzzlingly, has yielded few documented instances of substantial gains to domain pretraining in spite of the fact that legal language is widely seen to be unique. We hypothesize that these existing results stem from the fact that existing legal NLP tasks are too easy and fail to meet conditions for when domain pretraining can help. To address this, we first present CaseHOLD (Case Holdings On Legal Decisions), a new dataset comprised of over 53,000+ multiple choice questions to identify the relevant holding of a cited case. This dataset presents a fundamental task to lawyers and is both legally meaningful and difficult from an NLP perspective (F1 of 0.4 with a BiLSTM baseline). Second, we assess performance gains on CaseHOLD and existing legal NLP datasets. While a Transformer architecture (BERT) pretrained on a general corpus (Google Books and Wikipedia) improves performance, domain pretraining (using corpus of approximately 3.5M decisions across all courts in the U.S. that is larger than BERT's) with a custom legal vocabulary exhibits the most substantial performance gains with CaseHOLD (gain of 7.2% on F1, representing a 12% improvement on BERT) and consistent performance gains across two other legal tasks. Third, we show that domain pretraining may be warranted when the task exhibits sufficient similarity to the pretraining corpus: the level of performance increase in three legal tasks was directly tied to the domain specificity of the task. Our findings inform when researchers should engage resource-intensive pretraining and show that Transformer-based architectures, too, learn embeddings suggestive of distinct legal language., Comment: ICAIL 2021. Code & data available at https://github.com/reglab/casehold

Longitudinal studies are vital to understanding dynamic changes of the planet, but labels (e.g., buildings, facilities, roads) are often available only for a single point in time. We propose a general model, Temporal Cluster Matching (TCM), for detecting building changes in time series of remotely sensed imagery when footprint labels are observed only once. The intuition behind the model is that the relationship between spectral values inside and outside of building's footprint will change when a building is constructed (or demolished). For instance, in rural settings, the pre-construction area may look similar to the surrounding environment until the building is constructed. Similarly, in urban settings, the pre-construction areas will look different from the surrounding environment until construction. We further propose a heuristic method for selecting the parameters of our model which allows it to be applied in novel settings without requiring data labeling efforts (to fit the parameters). We apply our model over a dataset of poultry barns from 2016/2017 high-resolution aerial imagery in the Delmarva Peninsula and a dataset of solar farms from a 2020 mosaic of Sentinel 2 imagery in India. Our results show that our model performs as well when fit using the proposed heuristic as it does when fit with labeled data, and further, that supervised versions of our model perform the best among all the baselines we test against. Finally, we show that our proposed approach can act as an effective data augmentation strategy -- it enables researchers to augment existing structure footprint labels along the time dimension and thus use imagery from multiple points in time to train deep learning models. We show that this improves the spatial generalization of such models when evaluated on the same change detection task., Comment: Published in ACM COMPASS 2021

Computational methods that operate on three-dimensional molecular structure have the potential to solve important questions in biology and chemistry. In particular, deep neural networks have gained significant attention, but their widespread adoption in the biomolecular domain has been limited by a lack of either systematic performance benchmarks or a unified toolkit for interacting with molecular data. To address this, we present ATOM3D, a collection of both novel and existing benchmark datasets spanning several key classes of biomolecules. We implement several classes of three-dimensional molecular learning methods for each of these tasks and show that they consistently improve performance relative to methods based on one- and two-dimensional representations. The specific choice of architecture proves to be critical for performance, with three-dimensional convolutional networks excelling at tasks involving complex geometries, graph networks performing well on systems requiring detailed positional information, and the more recently developed equivariant networks showing significant promise. Our results indicate that many molecular problems stand to gain from three-dimensional molecular learning, and that there is potential for improvement on many tasks which remain underexplored. To lower the barrier to entry and facilitate further developments in the field, we also provide a comprehensive suite of tools for dataset processing, model training, and evaluation in our open-source atom3d Python package. All datasets are available for download from https://www.atom3d.ai ., Comment: NeurIPS 2021 Datasets and Benchmarks Track

The rise of machine learning (ML) has created an explosion in the potential strategies for using data to make scientific predictions. For physical scientists wishing to apply ML strategies to a particular domain, it can be difficult to assess in advance what strategy to adopt within a vast space of possibilities. Here we outline the results of an online community-powered effort to swarm search the space of ML strategies and develop algorithms for predicting atomic-pairwise nuclear magnetic resonance (NMR) properties in molecules. Using an open-source dataset, we worked with Kaggle to design and host a 3-month competition which received 47,800 ML model predictions from 2,700 teams in 84 countries. Within 3 weeks, the Kaggle community produced models with comparable accuracy to our best previously published "in-house" efforts. A meta-ensemble model constructed as a linear combination of the top predictions has a prediction accuracy which exceeds that of any individual model, 7-19x better than our previous state-of-the-art. The results highlight the potential of transformer architectures for predicting quantum mechanical (QM) molecular properties.

We present a neural network architecture that is fully equivariant with respect to transformations under the Lorentz group, a fundamental symmetry of space and time in physics. The architecture is based on the theory of the finite-dimensional representations of the Lorentz group and the equivariant nonlinearity involves the tensor product. For classification tasks in particle physics, we demonstrate that such an equivariant architecture leads to drastically simpler models that have relatively few learnable parameters and are much more physically interpretable than leading approaches that use CNNs and point cloud approaches. The competitive performance of the network is demonstrated on a public classification dataset [27] for tagging top quark decays given energy-momenta of jet constituents produced in proton-proton collisions.

We propose Cormorant, a rotationally covariant neural network architecture for learning the behavior and properties of complex many-body physical systems. We apply these networks to molecular systems with two goals: learning atomic potential energy surfaces for use in Molecular Dynamics simulations, and learning ground state properties of molecules calculated by Density Functional Theory. Some of the key features of our network are that (a) each neuron explicitly corresponds to a subset of atoms; (b) the activation of each neuron is covariant to rotations, ensuring that overall the network is fully rotationally invariant. Furthermore, the non-linearity in our network is based upon tensor products and the Clebsch-Gordan decomposition, allowing the network to operate entirely in Fourier space. Cormorant significantly outperforms competing algorithms in learning molecular Potential Energy Surfaces from conformational geometries in the MD-17 dataset, and is competitive with other methods at learning geometric, energetic, electronic, and thermodynamic properties of molecules on the GDB-9 dataset.

A Bose condensate subject to a periodic modulation of the two-body interactions was recently observed to emit matter-wave jets resembling "fireworks" [Nature 551, 356(2017)]. In this paper, combining experiment with numerical simulation, we demonstrate that these "Bose fireworks" represent a late stage in a complex time evolution of the driven condensate. We identify a "density wave" stage which precedes jet emission and results from interference of matterwaves. The density waves self-organize and self-amplify without the breaking of long range translational symmetry. Importantly, this density wave structure deterministically establishes the template for the subsequent patterns of the emitted jets. Our simulations, in good agreement with experiment, also address the apparent asymmetry in the jet pattern and show it is fully consistent with momentum conservation.

We study fermion-parity-changing quantum phase transitions (QPTs) in platform Josephson junctions. These QPTs, associated with zero-energy bound states, are rather widely observed experimentally. They emerge from numerical calculations frequently without detailed microscopic insight. Importantly, they may incorrectly lend support to claims for the observations of Majorana zero modes. In this paper we present a fully consistent solution of the Bogoliubov-de Gennes equations for a multi-component Josephson junction. This provides insights into the origin of the QPTs. It also makes it possible to assess the standard self energy approximations which are widely used to understand proximity coupling in topological systems. The junctions we consider are complex and chosen to mirror experiments. Our full proximity calculations associate the mechanism behind the QPT as deriving from a spatially extended, proximity-induced magnetic "defect". This defect arises because of the insulating region which effects a local reorganization of the bulk magnetization in the proximitized superconductor. Our results suggest more generally that QPTs in Josephson junctions generally do not require the existence of spin-orbit coupling and should not be confused with, nor are they indicators of, Majorana physics., Comment: 15 pages, 8 figures

We demonstrate a density-dependent gauge field, induced by atomic interactions, for quantum gases. The gauge field results from the synchronous coupling between the interactions and micromotion of the atoms in a modulated two-dimensional optical lattice. As a first step, we show that a coherent shaking of the lattice in two directions can couple the momentum and interactions of atoms and break the four-fold symmetry of the lattice. We then create a full interaction-induced gauge field by modulating the interaction strength in synchrony with the lattice shaking. When a condensate is loaded into this shaken lattice, the gauge field acts to preferentially prepare the system in different quasimomentum ground states depending on the modulation phase. We envision that these interaction-induced fields, created by fine control of micromotion, will provide a stepping stone to model new quantum phenomena within and beyond condensed matter physics.

Most existing neural networks for learning graphs address permutation invariance by conceiving of the network as a message passing scheme, where each node sums the feature vectors coming from its neighbors. We argue that this imposes a limitation on their representation power, and instead propose a new general architecture for representing objects consisting of a hierarchy of parts, which we call Covariant Compositional Networks (CCNs). Here, covariance means that the activation of each neuron must transform in a specific way under permutations, similarly to steerability in CNNs. We achieve covariance by making each activation transform according to a tensor representation of the permutation group, and derive the corresponding tensor aggregation rules that each neuron must implement. Experiments show that CCNs can outperform competing methods on standard graph learning benchmarks.

Topological- and strongly-correlated- materials are exciting frontiers in condensed matter physics, married prominently in studies of the fractional quantum hall effect [1]. There is an active effort to develop synthetic materials where the microscopic dynamics and ordering arising from the interplay of topology and interaction may be directly explored. In this work we demonstrate a novel architecture for exploration of topological matter constructed from tunnel-coupled, time-reversalbroken microwave cavities that are both low loss and compatible with Josephson junction-mediated interactions [2]. Following our proposed protocol [3] we implement a square lattice Hofstadter model at a quarter flux per plaquette ({\alpha} = 1/4), with time-reversal symmetry broken through the chiral Wannier-orbital of resonators coupled to Yttrium-Iron-Garnet spheres. We demonstrate site-resolved spectroscopy of the lattice, time-resolved dynamics of its edge channels, and a direct measurement of the dispersion of the edge channels. Finally, we demonstrate the flexibility of the approach by erecting a tunnel barrier investigating dynamics across it. With the introduction of Josephson-junctions to mediate interactions between photons, this platform is poised to explore strongly correlated topological quantum science for the first time in a synthetic system., Comment: 11 pages, 9 Figures

In this paper we demonstrate the necessity of including the generally omitted collective mode contributions in calculations of the Meissner effect for non-uniform superconductors. We consider superconducting pairing with non-zero center of mass momentum, as is relevant to high transition temperature cuprates, cold atoms, and quantum chromodynamic superconductors. For the concrete example of the Fulde-Ferrell phase we present a quantitative calculation of the superfluid density, showing the collective mode contributions are not only appreciable but that they derive from the amplitude mode of the order parameter. This latter mode (related to the Higgs mode in a charged system) is generally viewed as being invisible in conventional superconductors. However, our analysis shows that it is extremely important in pair-density wave type superconductors, where it destroys superfluidity well before the mean-field order parameter vanishes.

Optical control and manipulation of cold atoms has become an important topic in condensed matter. Widely employed are optical lattice shaking experiments which allow the introduction of artificial gauge fields, the design of topological bandstructures, and more general probing of quantum critical phenomena. Here we develop new numerical methods to simulate these periodically driven systems by implementing lattice shaking directly. As a result we avoid the usual assumptions associated with a simplified picture based on Floquet dynamics. A demonstrable success of our approach is that it yields quantitative agreement with experiment, including Kibble-Zurek scaling. Importantly, we argue that because their dynamics corresponds to an effective non-linear Schr\"{o}dinger equation, these particular superfluid studies present a unique opportunity to address how general Floquet band engineering is affected by interactions. In particular, interactions cause instabilities at which the behavior of the system changes dramatically.

We show that topological phases should be realizable in readily available and well studied heterostructures. In particular we identify a new class of topological materials which are well known in spintronics: helical ferromagnet-superconducting junctions. We note that almost all previous work on topological heterostructures has focused on creating Majorana modes at the proximity interface in effectively two-dimensional or one-dimensional systems. The particular heterostructures we address exhibit finite range proximity effects leading to nodal superconductors with Majorana modes localized well away from this interface. To show this, we implement a Bogoliubov-de Gennes (BdG) proximity numerical scheme, which importantly, involves two finite dimensions in a three dimensional junction. Incorporating this level of numerical complexity serves to distinguish ours from alternative numerical BdG approaches which are limited by generally assuming translational invariance or periodic boundary conditions along multiple directions. With this access to the edges, we are then able to illustrate in a concrete fashion the wavefunctions of Majorana zero modes, and, moreover, address finite size effects. In the process we establish consistency with a simple analytical model.

We present a scalable architecture for the exploration of interacting topological phases of photons in arrays of microwave cavities, using established techniques from cavity and circuit quantum electrodynamics. A time-reversal symmetry breaking (non-reciprocal) flux is induced by coupling the microwave cavities to ferrites, allowing for the production of a variety of topological band structures including the $\alpha=1/4$ Hofstadter model. Effective photon-photon interactions are included by coupling the cavities to superconducting qubits, and are sufficient to produce a $\nu=1/2$ bosonic Laughlin puddle. We demonstrate by exact diagonalization that this architecture is robust to experimentally achievable levels of disorder. These advances provide an exciting opportunity to employ the quantum circuit toolkit for the exploration of strongly interacting topological materials.

The nature of dark matter is a longstanding enigma of physics; it may consist of particles beyond the Standard Model that are still elusive to experiments. Among indirect search techniques, which look for stable products from the annihilation or decay of dark matter particles, or from axions coupling to high-energy photons, observations of the $\gamma$-ray sky have come to prominence over the last few years, because of the excellent sensitivity of the Large Area Telescope (LAT) on the Fermi Gamma-ray Space Telescope mission. The LAT energy range from 20 MeV to above 300 GeV is particularly well suited for searching for products of the interactions of dark matter particles. In this report we describe methods used to search for evidence of dark matter with the LAT, and review the status of searches performed with up to six years of LAT data. We also discuss the factors that determine the sensitivities of these searches, including the magnitudes of the signals and the relevant backgrounds, considering both statistical and systematic uncertainties. We project the expected sensitivities of each search method for 10 and 15 years of LAT data taking. In particular, we find that the sensitivity of searches targeting dwarf galaxies, which provide the best limits currently, will improve faster than the square root of observing time. Current LAT limits for dwarf galaxies using six years of data reach the thermal relic level for masses up to 120 GeV for the $b\bar{b}$ annihilation channel for reasonable dark matter density profiles. With projected discoveries of additional dwarfs, these limits could extend to about 250 GeV. With as much as 15 years of LAT data these searches would be sensitive to dark matter annihilations at the thermal relic cross section for masses to greater than 400 GeV (200 GeV) in the $b\bar{b}$ ($\tau^+ \tau^-$) annihilation channels., Comment: Updated with a few additional and corrected references; otherwise, text is identical to previous version. Submitted on behalf of the Fermi-LAT collaboration. Accepted for publication in Physics Reports, 59 pages, 34 figures; corresponding author: Eric Charles (echarles@slac.stanford.edu)

We study a trapped two-dimensional spin-imbalanced Fermi gas over a range of temperatures. In the moderate temperature regime, associated with current experiments, we find reasonable semi-quantitative agreement with the measured density profiles as functions of varying spin imbalance and interaction strength. Our calculations show that, in contrast to the three-dimensional case, the phase separation which appears as a spin balanced core, can be associated with non-condensed fermion pairs. We present predictions at lower temperatures where a quasi-condensate will first appear, based on the pair momentum distribution and following the protocols of Jochim and collaborators. While these profiles also indicate phase separation, they exhibit distinctive features which may aid in identifying the condensation regime., Comment: 4 pages, 4 figure

The search for new significant peaks over a energy spectrum often involves a statistical multiple hypothesis testing problem. Separate tests of hypothesis are conducted at different locations producing an ensemble of local p-values, the smallest of which is reported as evidence for the new resonance. Unfortunately, controlling the false detection rate (type I error rate) of such procedures may lead to excessively stringent acceptance criteria. In the recent physics literature, two promising statistical tools have been proposed to overcome these limitations. In 2005, a method to "find needles in haystacks" was introduced by Pilla et al. [1], and a second method was later proposed by Gross and Vitells [2] in the context of the "look elsewhere effect" and trial factors. We show that, for relatively small sample sizes, the former leads to an artificial inflation of statistical power that stems from an increase in the false detection rate, whereas the two methods exhibit similar performance for large sample sizes. We apply the methods to realistic simulations of the Fermi Large Area Telescope data, in particular the search for dark matter annihilation lines. Further, we discuss the counter-intutive scenario where the look-elsewhere corrections are more conservative than much more computationally efficient corrections for multiple hypothesis testing. Finally, we provide general guidelines for navigating the tradeoffs between statistical and computational efficiency when selecting a statistical procedure for signal detection.

In this paper we show how to redress a shortcoming of the path integral scheme for fermionic superfluids and superconductors. This approach is built around a simultaneous calculation of electrodynamics and thermodynamics. An important sum rule, the compressibility sum rule, fails to be satisfied in the usual calculation of the electromagnetic and thermodynamic response at the Gaussian fluctuation level. Here we present a path integral scheme to address this inconsistency. Specifically, at the leading order we argue that the superconducting gap should be calculated using a different saddle point condition modified by the presence of an external vector potential. This leads to the well known gauge-invariant BCS electrodynamic response and is associated with the usual (mean field) expression for thermodynamics. In this way the compressibility sum rule is satisfied at the BCS level. Moreover, this scheme can be readily extended to address arbitrary higher order fluctuation theories. At any level this approach will lead to a gauge invariant and compressibility sum rule consistent treatment of electrodynamics and thermodynamics., Comment: Comments welcome. Submitted directly to Phys. Rev. B Rapid Communications

We present a general diagrammatic theory for determining consistent electromagnetic response functions in strongly correlated fermionic superfluids. The general treatment of correlations beyond BCS theory requires a new theoretical formalism not contained in the current literature. Among concrete examples are a rather extensive class of theoretical models which incorporate BCS-BEC crossover as applied to the ultra cold Fermi gases, along with theories specifically associated with the high-$T_c$ cuprates. The challenge is to maintain gauge invariance, while simultaneously incorporating additional self-energy terms arising from strong correlation effects. Central to our approach is the application of the Ward-Takahashi identity, which introduces collective mode contributions in the response functions and guarantees that the $f$-sum rule is satisfied. We outline a powerful and very general method to determine these collective modes in a manner compatible with gauge invariance. Finally, as an alternative approach, we contrast with the path integral formalism. Here, the calculation of gauge invariant response appears more straightforward. However, the collective modes introduced are essentially those of strict BCS theory, with no modification from correlation effects. Since the path integral scheme simultaneously addresses electrodynamics and thermodynamics, we emphasize that it should be subjected to a consistency test beyond gauge invariance, namely that of the compressibility sum-rule. We show how this sum-rule fails in the conventional path integral approach., Comment: 10 pages + 7 pages Supplementary Material

In this paper we follow the analysis and protocols of recent experiments, combined with simple theory, to arrive at a physical understanding of quasi-condensation in two dimensional Fermi gases. We find that quasi-condensation mirrors Berezinskii-Kosterlitz-Thouless behavior in many ways, including the emergence of a strong zero momentum peak in the pair momentum distribution. Importantly, the disappearance of this quasi-condensate occurs at a reasonably well defined crossover temperature. The resulting phase diagram, pair momentum distribution, and algebraic power law decay are compatible with recent experiments throughout the continuum from BEC to BCS.

The dwarf spheroidal satellite galaxies of the Milky Way are some of the most dark-matter-dominated objects known. Due to their proximity, high dark matter content, and lack of astrophysical backgrounds, dwarf spheroidal galaxies are widely considered to be among the most promising targets for the indirect detection of dark matter via gamma rays. Here we report on gamma-ray observations of Milky Way dwarf spheroidal satellite galaxies based on 6 years of Fermi Large Area Telescope data processed with the new Pass 8 reconstruction and event-level analysis. None of the dwarf galaxies are significantly detected in gamma rays, and we present upper limits on the dark matter annihilation cross section from a combined analysis of the 15 most promising dwarf galaxies. The constraints derived are among the strongest to date using gamma rays and lie below the canonical thermal relic cross section for WIMPs of mass $\lesssim 100~GeV$ annihilating via the $b \bar b$ and $\tau^{+}\tau^{-}$ channels., Comment: 8 pages, 2 figures; in Proceedings of the 34th International Cosmic Ray Conference (ICRC 2015), The Hague (The Netherlands)

We investigate the effects of topological order on the transition temperature, $T_c$, and response functions in fermionic superfluids with Rashba spin-orbit coupling and a transverse Zeeman field in three dimensions. Our calculations, relevant to the ultracold atomic Fermi gases, include fluctuations beyond mean-field theory and are compatible with $f$-sum rules. Reminiscent of the $p_x + i p_y$ superfluid, the topological phase is stabilized when driven away from the Bose-Einstein condensation and towards the BCS limit. Accordingly, while experimentally accessible, $T_c$ is significantly suppressed in a topological superfluid. Above $T_c$, the spin and density response functions provide signatures of topological phases via the recombination or amplification of frequency dependent peaks.

Fractional Chern insulators are the proposed phases of matter mimicking the physics of fractional quantum Hall states on a lattice without an overall magnetic field. The notion of Floquet fractional Chern insulators refers to the potential possibilities to generate the underlying topological bandstructure by means of Floquet engineering. In these schemes, a highly controllable and strongly interacting system is periodically driven by an external force at a frequency such that double tunneling events during one forcing period become important and contribute to shaping the required effective energy bands. We show that in the described circumstances it is necessary to take into account also third order processes combining two tunneling events with interactions. Referring to the obtained contributions as micromotion-induced interactions, we find that those interactions tend to have a negative impact on the stability of of fractional Chern insulating phases and discuss implications for future experiments., Comment: 13 pages, 7 figures

We derive expressions for spin and density correlation functions in the (greatly enhanced) pseudogap phase of spin-orbit coupled Fermi superfluids. Density-density correlation functions are found to be relatively insensitive to the presence of these Rashba effects. To arrive at spin-spin correlation functions we derive new $f$-sum rules, valid even in the absence of a spin conservation law. Our spin-spin correlation functions are shown to be fully consistent with these $f$-sum rules. Importantly, they provide a clear signature of the Rashba band-structure and separately help to establish the presence of a pseudogap., Comment: 5 pages, 2 figures, with 5 page supplement

The joint likelihood is a simple extension of the standard likelihood formalism that enables the estimation of common parameters across disjoint datasets. Joining the likelihood, rather than the data itself, means nuisance parameters can be dealt with independently. Application of this technique, particularly to Fermi-LAT dwarf spheroidal analyses, has already been met with great success. We present a description of the method's general implementation along with a toy Monte-Carlo study of its properties and limitations., Comment: 2014 Fermi Symposium proceedings - eConf C14102.1

We present experimental evidence showing that an interacting Bose condensate in a shaken optical lattice develops a roton-maxon excitation spectrum, a feature normally associated with superfluid helium. The roton-maxon feature originates from the double-well dispersion in the shaken lattice, and can be controlled by both the atomic interaction and the lattice shaking amplitude. We determine the excitation spectrum using Bragg spectroscopy and measure the critical velocity by dragging a weak speckle potential through the condensate - both techniques are based on a digital micromirror device. Our dispersion measurements are in good agreement with a modified-Bogoliubov model., Comment: 9 pages, 7 figures

We derive an effective ring model in momentum space for trapped bosons with synthetic spin-orbit coupling. This effective model is characterized by a peculiar form of the inter particle interactions, which is crucially modified by the external confinement. The ring model allows for an intuitive understanding of the phase diagram of trapped condensates with isotropic spin-orbit coupling, and in particular for the existence of skyrmion lattice phases. The model, which may be generally applied for spinor condensates of arbitrary spin and spin-dependent interactions, is illustrated for the particular cases of spin-1/2 and spin-1 condensates.

We study the ground state phases of a rotating two-component, or binary Bose-Einstein condensate, wherein one component possesses a large magnetic dipole moment. A variety of non-trivial phases emerge in this system, including a half-quantum vortex (HQV) chain phase and a HQV molecule phase, where HQVs of opposite charge bind at short distances. We attribute the emergence of these phases to the development of a minimum in the adiabatic HQV interaction potential, which we calculate explicitly. We thus show that the presence of dipolar interactions in this system leads to a rich phase diagram, and the formation of HQV molecules., Comment: 4+ pages, 4 figures

Bose-Einstein condensates (BECs) of positronium (Ps) have been of experimental and theoretical interest due to their potential application as the gain medium of a $\gamma$-ray laser. Ps BECs are intrinsically spinor due to the presence of ortho-positronium (o-Ps) and para-positronium (p-Ps), whose annihilation lifetimes differ by three orders of magnitude. In this paper, we study the spinor dynamics and annihilation processes in the p-Ps/o-Ps system using both solutions of the time-dependent Gross-Pitaevskii equations and a semiclassical rate-equation approach. The spinor interactions have an $O(4)$ symmetry which is broken to $SO(3)$ by an internal energy difference between o-Ps and p-Ps. For an initially unpolarized condensate, there is a threshold density of $\approx 10^{19}$ cm$^{-3}$ at which spin mixing between o-Ps and p-Ps occurs. Beyond this threshold, there are unstable spatial modes accompanied by spin mixing. To ensure a high production yield above the critical density, a careful choice of external field must be made to avoid the spin mixing instability., Comment: 14 pages, 9 figures

We study the effects of dipolar interactions on a Bose-Einstein condensate with synthetically generated Rashba spin-orbit coupling. The dipolar interaction we consider includes terms that couple spin and orbital angular momentum in a way perfectly congruent with the single-particle Rashba coupling. We show that this internal spin-orbit coupling plays a crucial role in the rich ground-state phase diagram of the trapped condensate. In particular, we predict the emergence of a thermodynamically stable ground state with a meron spin configuration., Comment: 4+ pages, 3 figures

We present a new technique for producing two- and three-dimensional Rashba-type spin-orbit couplings for ultracold atoms without involving light. The method relies on a sequence of pulsed inhomogeneous magnetic fields imprinting suitable phase gradients on the atoms. For sufficiently short pulse durations, the time-averaged Hamiltonian well approximates the Rashba Hamiltonian. Higher order corrections to the energy spectrum are calculated exactly for spin-1/2 and perturbatively for higher spins. The pulse sequence does not modify the form of rotationally symmetric atom-atom interactions. Finally, we present a straightforward implementation of this pulse sequence on an atom chip.

The dynamics, appearing after a quantum quench, of a trapped, spin-orbit coupled, dilute atomic gas is studied. The characteristics of the evolution is greatly influenced by the symmetries of the system, and we especially compare evolution for an isotropic Rashba coupling and for an anisotropic spin-orbit coupling. As we make the spin-orbit coupling anisotropic, we break the rotational symmetry and the underlying classical model becomes chaotic; the quantum dynamics is affected accordingly. Within experimentally relevant time-scales and parameters, the system thermalizes in a quantum sense. The corresponding equilibration time is found to agree with the Ehrenfest time, i.e. we numerically verify a ~log(1/h) scaling. Upon thermalization, we find the equilibrated distributions show examples of quantum scars distinguished by accumulation of atomic density for certain energies. At shorter time-scales we discuss non-adiabatic effects deriving from the spin-orbit coupled induced Dirac point. In the vicinity of the Dirac point, spin fluctuations are large and, even at short times, a semi-classical analysis fails., Comment: 11 pages, 10 figures

We investigate the properties of an atom under the influence of a synthetic three-dimensional spin-orbit coupling (Weyl coupling) in the presence of a harmonic trap. The conservation of total angular momentum provides a numerically efficient scheme for finding the spectrum and eigenfunctions of the system. We show that at large spin-orbit coupling the system undergoes dimensional reduction from three to one dimension at low energies, and the spectrum is approximately Landau level-like. At high energies, the spectrum is approximately given by the three-dimensional isotropic harmonic oscillator. We explore the properties of the ground state in both position and momentum space. We find the ground state has spin textures with oscillations set by the spin-orbit length scale.

We describe a method for creating a three-dimensional analogue to Rashba spin-orbit coupling in systems of ultracold atoms. This laser induced coupling uses Raman transitions to link four internal atomic states with a tetrahedral geometry, and gives rise to a Dirac point that is robust against environmental perturbations. We present an exact result showing that such a spin-orbit coupling in a fermionic system always rise to a molecular bound state., Comment: 4 pages, 6 figures

We study interacting Rashba-Dresselhaus fermions in two spatial dimensions. First, we present a new exact solution to the two-particle pairing problem of spin-orbit-coupled fermions for arbitrary Rashba and Dresselhaus spin-orbit interactions. An exact molecular wave function and the Green function are explicitly derived along with the binding energy and the spectrum of the molecular state. In the second part, we consider a thermal Boltzmann gas of fermionic molecules and compute the time-of-flight velocity and spin distributions for a single fermion in the gas. We show that the pairing signatures can be observed already in the first-moment expectation values, such as time-of-flight density and spin profiles., Comment: 12 pages, 8 figures; Fig. 1 fixed