Search

We consider the problem of optimal exchange which can be formulated as a kind of optimal transportation problem. The existence of an optimal solution and a duality theorem for the optimal exchange problem are proved in case of completely regular topological spaces. We show the connection between the problem of optimal exchange and the optimal transportation problem with density constraints. With the use of this connection we obtain a formula for the optimal value in the problem of optimal exchange.

Removing modeling constraints and unifying architectures across domains has been a key driver of the recent progress in training large multimodal models. However, most of these models still rely on many separately trained components such as modality-specific encoders and decoders. In this work, we further streamline joint generative modeling of images and text. We propose an autoregressive decoder-only transformer - JetFormer - which is trained to directly maximize the likelihood of raw data, without relying on any separately pretrained components, and can understand and generate both text and images. Specifically, we leverage a normalizing flow model to obtain a soft-token image representation that is jointly trained with an autoregressive multimodal transformer. The normalizing flow model serves as both an image encoder for perception tasks and an image decoder for image generation tasks during inference. JetFormer achieves text-to-image generation quality competitive with recent VQ-VAE- and VAE-based baselines. These baselines rely on pretrained image autoencoders, which are trained with a complex mixture of losses, including perceptual ones. At the same time, JetFormer demonstrates robust image understanding capabilities. To the best of our knowledge, JetFormer is the first model that is capable of generating high-fidelity images and producing strong log-likelihood bounds.

Triangular lattice antiferromagnets are prototypes for frustrated magnetism and may potentially realize novel quantum magnetic states such as a quantum spin liquid ground state. A recent work suggests NdTa$_7$O$_{19}$ with rare-earth triangular lattice is a quantum spin liquid candidate and highlights the large family of rare-earth heptatantalates as a framework for quantum magnetism investigation. In this paper, we report the structural and magnetic characterization of CeTa$_7$O$_{19}$ and YbTa$_7$O$_{19}$. Both compounds are isostructural to NdTa$_7$O$_{19}$ with no detectable structural disorder. For CeTa$_7$O$_{19}$, the crystal field energy levels and parameters are determined by inelastic neutron scattering measurements. Based on the crystal field result, the magnetic susceptibility data could be well fitted and explained, which reveals that CeTa$_7$O$_{19}$ is a highly anisotropic Ising triangular-lattice antiferromagnet ($g_z$/$g_{xy}$$\sim$3) with very weak exchange interaction (J$\sim$0.22~K). For YbTa$_7$O$_{19}$, millimeter sized single crystals could be grown. The anisotropic magnetization and electron spin resonance data show that YbTa$_7$O$_{19}$ has a contrasting in-plane magnetic anisotropy with $g_z$/$g_{xy}$$\sim$0.67 similar as that of YbMgGaO$_4$. The above results indicate that CeTa$_7$O$_{19}$ and YbTa$_7$O$_{19}$ with pseudospin-1/2 ground states might either be quantum spin liquid candidate materials or find applications in adiabatic demagnetization refrigeration due to the weak exchange interaction., Comment: 10 pages, 7 figures

In this paper, we study new extensions of the functional Blaschke-Santalo inequalities, and explore applications of such new inequalities beyond the classical setting of the standard Gaussian measure.

One of the most important issues in modern condensed matter physics is the realization of fractionalized excitations, such as the Majorana excitations in the Kitaev quantum spin liquid. To this aim, the 3d-based Kitaev material Na2Co2TeO6 is a promising candidate whose magnetic phase diagram of B // a* contains a field-induced intermediate magnetically disordered phase within 7.5 T < |B| < 10 T. The experimental observations, including the restoration of the crystalline point group symmetry in the angle-dependent torque and the coexisting magnon excitations and spinon-continuum in the inelastic neutron scattering spectrum, provide strong evidence that this disordered phase is a field induced quantum spin liquid with partially polarized spins. Our variational Monte Carlo simulation with the effective K-J1-{\Gamma}-{\Gamma}'-J3 model reproduces the experimental data and further supports this conclusion.

PaliGemma is an open Vision-Language Model (VLM) that is based on the SigLIP-So400m vision encoder and the Gemma-2B language model. It is trained to be a versatile and broadly knowledgeable base model that is effective to transfer. It achieves strong performance on a wide variety of open-world tasks. We evaluate PaliGemma on almost 40 diverse tasks including standard VLM benchmarks, but also more specialized tasks such as remote-sensing and segmentation., Comment: v2 adds Appendix H and I and a few citations

Building generalized models that can solve many computer vision tasks simultaneously is an intriguing direction. Recent works have shown image itself can be used as a natural interface for general-purpose visual perception and demonstrated inspiring results. In this paper, we explore diffusion-based vision generalists, where we unify different types of dense prediction tasks as conditional image generation and re-purpose pre-trained diffusion models for it. However, directly applying off-the-shelf latent diffusion models leads to a quantization issue. Thus, we propose to perform diffusion in pixel space and provide a recipe for finetuning pre-trained text-to-image diffusion models for dense vision tasks. In experiments, we evaluate our method on four different types of tasks and show competitive performance to the other vision generalists., Comment: Published at CVPR 2024 as a workshop paper

In this study, a novel material from the rare-earth based breathing pyrochlore family, Ba3Tm2Zn5O11, was successfully synthesized. Powder X-ray diffraction and high-resolution powder neutron diffraction confirmed phase purity and crystal structure, while thermogravimetric analysis revealed incongruent melting behavior compared to its counterpart, Ba3Yb2Zn5O11. High-quality single crystals of Ba3Tm2Zn5O11 were grown using the traveling solvent floating zone technique and assessed using Laue diffractometer and single crystal X-ray diffraction. Thermodynamic characterization indicated paramagnetic behavior down to 0.05 K, and inelastic neutron scattering measurements identified distinct crystal electric field bands. Additional low-energy excitation studies on single crystals revealed dispersionless bands at 0.8 and 1 meV. Computed phonon dispersions from first principles calculations ruled out phonons as the source of these modes, further illustrating the puzzling and unique properties of Ba3Tm2Zn5O11.

Geometrically frustrated magnetic systems, such as those based on the Shastry-Sutherland lattice, offer a rich playground for exploring unconventional magnetic states. The delicate balance between competing interactions in these systems leads to the emergence of novel phases. We present the characterization of Er2Be2GeO7, an SSL compound with Er3+ ions forming orthogonal dimers separated by non-magnetic layers. Neutron scattering reveals an antiferromagnetic dimer structure at zero field, typical of Ising spins on that lattice and consistent with the anisotropic magnetization observed. However, magnetization measurements exhibit fractional plateaus at 1/4 and 1/2 of saturation, by contrast to the expected 1/3 plateau of the SSL Ising model. We show that this behaviour requires spatially anisotropic interactions, and, using a tensor network approach, we show that all properties, including thermodynamics, can be quantitatively explained by an Anisotropic Shastry-Sutherland Ising Model whose microscopic origin can be traced back to Dzyaloshinskii-Moriya interactions. This study highlights Er2Be2GeO7 as a promising platform for investigating exotic magnetic phenomena and potential applications in spintronics.

Two-dimensional (2D) layered materials, whether in bulk form or reduced to just a single layer, have potential applications in spintronics and capacity for advanced quantum phenomena. A prerequisite for harnessing these opportunities lies in gaining a comprehensive understanding of the spin behavior in 2D materials. The low dimensionality motivates an understanding of the spin correlations over a wide length scale, from local to long range order. In this context, we focus on the magnetism in bulk \MPSe ~and \MPS, 2D layered van der Waals antiferromagnetic semiconductors. These materials have similar honeycomb Mn layers and magnetic ordering temperatures, but distinct spin orientations and exchange interactions. We utilize neutron scattering to gain deeper insights into the local magnetic structures and spin correlations in the paramagnetic and ordered phases by systematically investigating a MnPS$_{x}$Se$_{3-x}$ ($x$ = 0, 1, 1.5, 2, 3) series of powder samples using total neutron scattering measurements. By employing magnetic pair distribution function (mPDF) analysis, we unraveled the short-range magnetic correlations in these materials and explored how the non-magnetic anion S/Se mixing impacts the magnetic correlations. The results reveal that the magnetism can be gradually tuned through alteration of the non-magnetic S/Se content, which tunes the atomic structure. The change in magnetic structure is also accompanied by a control of the magnetic correlation length within the 2D honeycomb layers. Complimentary inelastic neutron scattering measurements allowed a quantification of the change in the magnetic exchange interactions for the series and further highlighted the gradual evolution of spin interactions in the series MnPS$_{x}$Se$_{3-x}$.

Recent advances in the study of cobaltites have unveiled their potential as a promising platform for realizing Kitaev physics in honeycomb systems and the Ising model in weakly coupled chain materials. In this manuscript, we explore the magnetic properties of pyroxene SrCoGe$_2$O$_6$ using a combination of neutron scattering, {\it ab initio} methods, and linear spin-wave theory. Through careful examination of inelastic neutron scattering powder spectra, we propose a modified Kitaev model to accurately describe the twisted chains of edge-sharing octahedra surrounding Co$^{2+}$ ions. The extended Kitaev-Heisenberg model, including a significant anisotropic bond-dependent exchange term with $K/|J|=0.96$, is identified as the key descriptor of the magnetic interactions in SrCoGe$_2$O$_6$. Furthermore, our heat capacity measurements reveal an effect of an external magnetic field (approximately 13~T) which shifts the system from a fragile antiferromagnetic ordering with $T_{\mathrm{N}}=9$~K to a field-induced state. We argue that pyroxenes, particularly those modified by substituting Ge with Si and its less extended $p$ orbitals, emerge as a novel platform for the Kitaev model. This opens up possibilities for advancing our understanding of Kitaev physics., Comment: 11 + 2 pages, 7 + 3 figures

The two-dimensional (2D) material Cr$_2$Ge$_2$Te$_6$ is a member of the class of insulating van der Waals magnets. Here, using high resolution angle-resolved photoemission spectroscopy in a detailed temperature dependence study, we identify a clear response of the electronic structure to a dimensional crossover in the form of two distinct temperature scales marking onsets of modifications in the electronic structure. Specifically, we observe Te $p$-orbital-dominated bands to undergo changes at the Curie transition temperature T$_C$ while the Cr $d$-orbital-dominated bands begin evolving at a higher temperature scale. Combined with neutron scattering, density functional theory calculations, and Monte Carlo simulations, we find that the electronic system can be consistently understood to respond sequentially to the distinct temperatures at which in-plane and out-of-plane spin correlations exceed a characteristic length scale. Our findings reveal the sensitivity of the orbital-selective electronic structure for probing the dynamical evolution of local moment correlations in vdW insulating magnets., Comment: PRB, in press

In this survey paper we present classical and recent results relating the auction design and the optimal transportation theory. In particular, we discuss in details the seminal result of Daskalakis, Deckelbaum and Tzamos \cite{DDT} about duality between auction design with $1$ bidder and the weak transportation problem. Later investigations revealed the connection of multi-bidder case to the Beckmann's transportation problem. In this paper we overview a number of works on related subjects (monopolist's problem, regularity issues, weak transportation, measure ordering etc.). In addition, we prove some new results on duality for unreduced mechanisms.

Kitaev interactions, arising from the interplay of frustration and bond anisotropy, can lead to strong quantum fluctuations and, in an ideal case, to a quantum-spin-liquid state. However, in many nonideal materials, spurious non-Kitaev interactions typically promote a zigzag antiferromagnetic order in the d-orbital transition metal compounds. By combining neutron scattering with muon-spin rotation and relaxation techniques, we provide new insights into the exotic properties of Na2Co2TeO6, a candidate Kitaev material. Below TN, the zero-field muon-spin relaxation rate becomes almost constant (at 0.45 us-1). We attribute this temperature-independent muon-spin relaxation rate to the strong quantum fluctuations, as well as to the frustrated Kitaev interactions. As the magnetic field increases, neutron scattering data indicate a much broader spin-wave-excitation gap at the K-point. Therefore, quantum fluctuations seem not only robust, but are even enhanced by the applied magnetic field. Our findings provide valuable hints for understanding the onset of the quantum-spin-liquid state in Kitaev materials., Comment: 28 pages, 11 figures, and 1 lable

This paper presents PaLI-3, a smaller, faster, and stronger vision language model (VLM) that compares favorably to similar models that are 10x larger. As part of arriving at this strong performance, we compare Vision Transformer (ViT) models pretrained using classification objectives to contrastively (SigLIP) pretrained ones. We find that, while slightly underperforming on standard image classification benchmarks, SigLIP-based PaLI shows superior performance across various multimodal benchmarks, especially on localization and visually-situated text understanding. We scale the SigLIP image encoder up to 2 billion parameters, and achieves a new state-of-the-art on multilingual cross-modal retrieval. We hope that PaLI-3, at only 5B parameters, rekindles research on fundamental pieces of complex VLMs, and could fuel a new generation of scaled-up models.

We have synthesized high-quality single crystal samples of the erbium-based triangular lattice compound Ba3Er(BO3)3. Thermal and magnetic measurements reveal large anisotropy and a possible phase transition at 100 mK. The low-temperature magnetic heat capacity can be understood from the two Wyckoff positions that the magnetic ions Er3+ occupy, which have distinct symmetry properties and crystal field environments. A point charge calculation for the crystal electric field levels is consistent with this understanding. Based on symmetry analysis and classical simulation, we argue that Ba3Er(BO3)3 realizes an interesting two-sublattice exchange physics, in which the honeycomb lattice spins develop ferromagnetic correlations due to the additional spins at the hexagon centers but eventually order antiferromagnetically. Additionally, our results suggest that quantum fluctuations need to be considered in order to fully explain the experimental observations.

The emergence of a quantum spin liquid (QSL), a state of matter that can result when electron spins are highly correlated but do not become ordered, has been the subject of a considerable body of research in condensed matter physics. Spin liquid states have been proposed as hosts for high-temperature superconductivity and can host topological properties with potential applications in quantum information science. The excitations of most quantum spin liquids are not conventional spin waves but rather quasiparticles known as spinons, whose existence is well established experimentally only in one-dimensional systems; the unambiguous experimental realization of QSL behavior in higher dimensions remains challenging. Here we investigate the novel compound YbZn2GaO5, which hosts an ideal triangular lattice of effective spin-1/2 moments with no inherent chemical disorder. Thermodynamic and inelastic neutron scattering (INS) measurements performed on high-quality single crystal samples of YbZn2GaO5 exclude the possibility of long-range magnetic ordering down to 60 mK, demonstrate a quadratic power law for the heat capacity and reveal a continuum of magnetic excitations in parts of the Brillouin zone. Both low-temperature thermodynamics and INS spectra suggest that YbZn2GaO5 is a U(1) Dirac QSL with gapless spinon excitations concentrated at certain points in the Brillouin zone, and additional features in INS are also consistent with theoretical expectations for a Dirac QSL on the triangular lattice.

We present the training recipe and results of scaling up PaLI-X, a multilingual vision and language model, both in terms of size of the components and the breadth of its training task mixture. Our model achieves new levels of performance on a wide-range of varied and complex tasks, including multiple image-based captioning and question-answering tasks, image-based document understanding and few-shot (in-context) learning, as well as object detection, video question answering, and video captioning. PaLI-X advances the state-of-the-art on most vision-and-language benchmarks considered (25+ of them). Finally, we observe emerging capabilities, such as complex counting and multilingual object detection, tasks that are not explicitly in the training mix.

Scaling laws have been recently employed to derive compute-optimal model size (number of parameters) for a given compute duration. We advance and refine such methods to infer compute-optimal model shapes, such as width and depth, and successfully implement this in vision transformers. Our shape-optimized vision transformer, SoViT, achieves results competitive with models that exceed twice its size, despite being pre-trained with an equivalent amount of compute. For example, SoViT-400m/14 achieves 90.3% fine-tuning accuracy on ILSRCV2012, surpassing the much larger ViT-g/14 and approaching ViT-G/14 under identical settings, with also less than half the inference cost. We conduct a thorough evaluation across multiple tasks, such as image classification, captioning, VQA and zero-shot transfer, demonstrating the effectiveness of our model across a broad range of domains and identifying limitations. Overall, our findings challenge the prevailing approach of blindly scaling up vision models and pave a path for a more informed scaling., Comment: 10 pages, 7 figures, 9 tables. Version 2: Layout fixes

The observation and description of collective excitations in solids is a fundamental issue when seeking to understand the physics of a many-body system. Analysis of these excitations is usually carried out by measuring the dynamical structure factor, S(Q, $\omega$), with inelastic neutron or x-ray scattering techniques and comparing this against a calculated dynamical model. Here, we develop an artificial intelligence framework which combines a neural network trained to mimic simulated data from a model Hamiltonian with automatic differentiation to recover unknown parameters from experimental data. We benchmark this approach on a Linear Spin Wave Theory (LSWT) simulator and advanced inelastic neutron scattering data from the square-lattice spin-1 antiferromagnet La$_2$NiO$_4$. We find that the model predicts the unknown parameters with excellent agreement relative to analytical fitting. In doing so, we illustrate the ability to build and train a differentiable model only once, which then can be applied in real-time to multi-dimensional scattering data, without the need for human-guided peak finding and fitting algorithms. This prototypical approach promises a new technology for this field to automatically detect and refine more advanced models for ordered quantum systems., Comment: 12 pages, 7 figures

There has been a recent explosion of computer vision models which perform many tasks and are composed of an image encoder (usually a ViT) and an autoregressive decoder (usually a Transformer). However, most of this work simply presents one system and its results, leaving many questions regarding design decisions and trade-offs of such systems unanswered. In this work, we aim to provide such answers. We take a close look at autoregressive decoders for multi-task learning in multimodal computer vision, including classification, captioning, visual question answering, and optical character recognition. Through extensive systematic experiments, we study the effects of task and data mixture, training and regularization hyperparameters, conditioning type and specificity, modality combination, and more. Importantly, we compare these to well-tuned single-task baselines to highlight the cost incurred by multi-tasking. A key finding is that a small decoder learned on top of a frozen pretrained encoder works surprisingly well. We call this setup locked-image tuning with decoder (LiT-decoder). It can be seen as teaching a decoder to interact with a pretrained vision model via natural language.

We propose a simple pairwise Sigmoid loss for Language-Image Pre-training (SigLIP). Unlike standard contrastive learning with softmax normalization, the sigmoid loss operates solely on image-text pairs and does not require a global view of the pairwise similarities for normalization. The sigmoid loss simultaneously allows further scaling up the batch size, while also performing better at smaller batch sizes. Combined with Locked-image Tuning, with only four TPUv4 chips, we train a SigLiT model that achieves 84.5% ImageNet zero-shot accuracy in two days. The disentanglement of the batch size from the loss further allows us to study the impact of examples vs pairs and negative to positive ratio. Finally, we push the batch size to the extreme, up to one million, and find that the benefits of growing batch size quickly diminish, with a more reasonable batch size of 32k being sufficient. We release our models at https://github.com/google-research/big_vision and hope our research motivates further explorations in improving the quality and efficiency of language-image pre-training., Comment: ICCV'23 Oral. arXiv v2: fix typo in pseudocode; v3: clarify t vs t' init; v4: add SigLIP Base, Large, Shape-Optimized 400M results. Models released at: https://github.com/google-research/big_vision. Xiaohua and Lucas contributed equally

We report inelastic neutron scattering measurements of magnetic excitations in YbMnSb2, a low-carrier-density Dirac semimetal in which the antiferromagnetic Mn layers are interleaved with Sb layers that host Dirac fermions. We observe a considerable broadening of spin waves, which is consistent with substantial spin fermion coupling. The spin wave damping, $\gamma$, in YbMnSb2 is roughly twice larger compared to that in a sister material, YbMnBi2, where an indication of a small damping consistent with theoretical analysis of the spin-fermion coupling was reported. The inter-plane interaction between the Mn layers in YbMnSb2 is also much stronger, suggesting that the interaction mechanism is rooted in the same spin-fermion coupling. Our results establish the systematics of spin-fermion interactions in layered magnetic Dirac materials.

Misalignment between model predictions and intended usage can be detrimental for the deployment of computer vision models. The issue is exacerbated when the task involves complex structured outputs, as it becomes harder to design procedures which address this misalignment. In natural language processing, this is often addressed using reinforcement learning techniques that align models with a task reward. We adopt this approach and show its surprising effectiveness across multiple computer vision tasks, such as object detection, panoptic segmentation, colorization and image captioning. We believe this approach has the potential to be widely useful for better aligning models with a diverse range of computer vision tasks., Comment: 11 pages

The scaling of Transformers has driven breakthrough capabilities for language models. At present, the largest large language models (LLMs) contain upwards of 100B parameters. Vision Transformers (ViT) have introduced the same architecture to image and video modelling, but these have not yet been successfully scaled to nearly the same degree; the largest dense ViT contains 4B parameters (Chen et al., 2022). We present a recipe for highly efficient and stable training of a 22B-parameter ViT (ViT-22B) and perform a wide variety of experiments on the resulting model. When evaluated on downstream tasks (often with a lightweight linear model on frozen features), ViT-22B demonstrates increasing performance with scale. We further observe other interesting benefits of scale, including an improved tradeoff between fairness and performance, state-of-the-art alignment to human visual perception in terms of shape/texture bias, and improved robustness. ViT-22B demonstrates the potential for "LLM-like" scaling in vision, and provides key steps towards getting there.

Lanthanides in the trivalent oxidation state are typically described using an ionic picture that leads to localized magnetic moments. The hierarchical energy scales associated with trivalent lanthanides produce desirable properties for e.g., molecular magnetism, quantum materials, and quantum transduction. Here, we show that this traditional ionic paradigm breaks down for praseodymium in the 4+ oxidation state. Synthetic, spectroscopic, and theoretical tools deployed on several solid-state Pr4+ oxides uncover the unusual participation of 4f orbitals in bonding and the anomalous hybridization of the 4f1 configuration with ligand valence electrons, analogous to transition metals. The resulting competition between crystal-field and spin-orbit-coupling interactions fundamentally transforms the spin-orbital magnetism of Pr4+, which departs from the Jeff =1/2 limit and resembles that of high-valent actinides. Our results show that Pr4+ ions are in a class on their own, where the hierarchy of single-ion energy scales can be tailored to explore new correlated phenomena in quantum materials., Comment: 9 pages, 4 figures, and SI (47 pages)

Vision Transformers convert images to sequences by slicing them into patches. The size of these patches controls a speed/accuracy tradeoff, with smaller patches leading to higher accuracy at greater computational cost, but changing the patch size typically requires retraining the model. In this paper, we demonstrate that simply randomizing the patch size at training time leads to a single set of weights that performs well across a wide range of patch sizes, making it possible to tailor the model to different compute budgets at deployment time. We extensively evaluate the resulting model, which we call FlexiViT, on a wide range of tasks, including classification, image-text retrieval, open-world detection, panoptic segmentation, and semantic segmentation, concluding that it usually matches, and sometimes outperforms, standard ViT models trained at a single patch size in an otherwise identical setup. Hence, FlexiViT training is a simple drop-in improvement for ViT that makes it easy to add compute-adaptive capabilities to most models relying on a ViT backbone architecture. Code and pre-trained models are available at https://github.com/google-research/big_vision, Comment: Code and pre-trained models available at https://github.com/google-research/big_vision. All authors made significant technical contributions. CVPR 2023

Effective scaling and a flexible task interface enable large language models to excel at many tasks. We present PaLI (Pathways Language and Image model), a model that extends this approach to the joint modeling of language and vision. PaLI generates text based on visual and textual inputs, and with this interface performs many vision, language, and multimodal tasks, in many languages. To train PaLI, we make use of large pre-trained encoder-decoder language models and Vision Transformers (ViTs). This allows us to capitalize on their existing capabilities and leverage the substantial cost of training them. We find that joint scaling of the vision and language components is important. Since existing Transformers for language are much larger than their vision counterparts, we train a large, 4-billion parameter ViT (ViT-e) to quantify the benefits from even larger-capacity vision models. To train PaLI, we create a large multilingual mix of pretraining tasks, based on a new image-text training set containing 10B images and texts in over 100 languages. PaLI achieves state-of-the-art in multiple vision and language tasks (such as captioning, visual question-answering, scene-text understanding), while retaining a simple, modular, and scalable design., Comment: ICLR 2023 (Notable-top-5%)

Spin and lattice are two fundamental degrees of freedom in a solid, and their fluctuations about the equilibrium values in a magnetic ordered crystalline lattice form quasiparticles termed magnons (spin waves) and phonons (lattice waves), respectively. In most materials with strong spin-lattice coupling (SLC), the interaction of spin and lattice induces energy gaps in the spin wave dispersion at the nominal intersections of magnon and phonon modes. Here we use neutron scattering to show that in the two-dimensional (2D) van der Waals honeycomb lattice ferromagnetic CrGeTe3, spin waves propagating within the 2D plane exhibit an anomalous dispersion, damping, and break-down of quasiparticle conservation, while magnons along the c axis behave as expected for a local moment ferromagnet. These results indicate the presence of dynamical SLC arising from the zero-temperature quantum fluctuations in CrGeTe3, suggesting that the observed in-plane spin waves are mixed spin and lattice quasiparticles fundamentally different from pure magnons and phonons., Comment: 12 pages, 4 figures; Nature Communication in press

Recently, Co-based honeycomb magnets have been proposed as promising candidate materials to host the Kitaev spin liquid state. One of the front-runners is BaCo$_2$(AsO$_4$)$_2$ (BCAO), where it was suggested that the exchange processes between Co$^{2+}$ ions via the surrounding edge-sharing oxygen octahedra could give rise to bond-dependent Kitaev interactions. In this work, we present and analyze comprehensive inelastic neutron scattering studies of BCAO with fields in the honeycomb plane. Combining the constraints from the magnon excitations in the high-field polarized state and the inelastic spin structure factor measured in zero magnetic field, we examine two leading theoretical models: the Kitaev-type \JKG model and the \XXZ model. We show that the existing experimental data can be consistently accounted for by the \XXZ model but not by the \JKG model, and we discuss the implications of these results for the realization of a spin liquid phase in BCAO and more generally for the realization of the Kitaev model in cobaltates., Comment: 15 pages, 10 figures

We introduce UViM, a unified approach capable of modeling a wide range of computer vision tasks. In contrast to previous models, UViM has the same functional form for all tasks; it requires no task-specific modifications which require extensive human expertise. The approach involves two components: (I) a base model (feed-forward) which is trained to directly predict raw vision outputs, guided by a learned discrete code and (II) a language model (autoregressive) that is trained to generate the guiding code. These components complement each other: the language model is well-suited to modeling structured interdependent data, while the base model is efficient at dealing with high-dimensional outputs. We demonstrate the effectiveness of UViM on three diverse and challenging vision tasks: panoptic segmentation, depth prediction and image colorization, where we achieve competitive and near state-of-the-art results. Our experimental results suggest that UViM is a promising candidate for a unified modeling approach in computer vision., Comment: 22 pages. Accepted at NeurIPS 2022

It is commonly accepted that the Vision Transformer model requires sophisticated regularization techniques to excel at ImageNet-1k scale data. Surprisingly, we find this is not the case and standard data augmentation is sufficient. This note presents a few minor modifications to the original Vision Transformer (ViT) vanilla training setting that dramatically improve the performance of plain ViT models. Notably, 90 epochs of training surpass 76% top-1 accuracy in under seven hours on a TPUv3-8, similar to the classic ResNet50 baseline, and 300 epochs of training reach 80% in less than one day., Comment: Code available at https://github.com/google-research/big_vision

We consider the problem of revenue-maximizing Bayesian auction design with several bidders having independent private values over several items. We show that it can be reduced to the problem of continuous optimal transportation introduced by Beckmann (1952) where the optimal transportation flow generalizes the concept of ironed virtual valuations to the multi-item setting. We establish the strong duality between the two problems and the existence of solutions. The results rely on insights from majorization and optimal transportation theories and on the characterization of feasible interim mechanisms by Hart and Reny (2015).

Fe3-xGeTe2 is a layered magnetic van der Waals material of interest for both fundamental and applied research. Despite the observation of intriguing physical properties, open questions exist even on the basic features related to magnetism: is it a simple ferromagnet or are there antiferromagnetic regimes and are the moments local or itinerant. Here, we demonstrate that antiferromagnetic spin fluctuations coexist with the ferromagnetism through comprehensive elastic and inelastic neutron scattering and thermodynamic measurements. Our realistic dynamical mean-field theory calculations reveal that the competing magnetic fluctuations are driven by an orbital selective Mott transition, where only the plane-perpendicular a1g orbital of the Fe(3d) manifold remains itinerant. Our results highlight the multi-orbital character in Fe3-xGeTe2 that supports a rare coexistence of local and itinerant physics within this material., Comment: 16 pages, 3 main figures

This paper presents contrastive-tuning, a simple method employing contrastive training to align image and text models while still taking advantage of their pre-training. In our empirical study we find that locked pre-trained image models with unlocked text models work best. We call this instance of contrastive-tuning "Locked-image Tuning" (LiT), which just teaches a text model to read out good representations from a pre-trained image model for new tasks. A LiT model gains the capability of zero-shot transfer to new vision tasks, such as image classification or retrieval. The proposed LiT is widely applicable; it works reliably with multiple pre-training methods (supervised and unsupervised) and across diverse architectures (ResNet, Vision Transformers and MLP-Mixer) using three different image-text datasets. With the transformer-based pre-trained ViT-g/14 model, the LiT model achieves 85.2% zero-shot transfer accuracy on the ImageNet test set, and 82.5% on the challenging out-of-distribution ObjectNet test set., Comment: Xiaohua, Xiao, Basil, Andreas and Lucas contributed equally; CVPR 2022