Search

Magnetization noise measurements on the spin ice Dy${}_2$Ti${}_2$O${}_7$ have revealed a remarkable `pink noise' power spectrum $S(f,T)$ below 4 K, including evidence of magnetic monopole excitations diffusing in a fractal landscape. However, at higher temperatures, the reported values of the anomalous exponent $b(T)$ describing the high frequency tail of $S(f,T)$ are not easy to reconcile with other results in the literature, which generally suggest significantly smaller deviations from the Brownian motion value of $b=2$, that become negligible above $T=20$ K. We accurately estimate $b(T)$ at temperatures between 2~K and 20~K, using a.c. susceptibility measurements that, crucially, stretch up to the relatively high frequency of $f = 10^6$ Hz. We show that previous noise measurements underestimate $b(T)$ and we suggest reasons for this. Our results establish deviations in $b(T)$ from $b=2$ up to about 20 K. However studies on different samples confirms that $b(T)$ is sample dependent: the details of this dependence agree in part, though not completely, with previous studies of the effect of crystal defects on monopole population and diffusion. Our results establish the form of $b(T)$ which characterises the subtle, and evolving, nature of monopole diffusion in the dense Coulomb fluid, a highly correlated state, where several dynamical processes combine. They do not rule out the importance of a fractal landscape picture emerging at lower temperatures where the monopole gas is dilute.

The symbiosis of strong interactions, flat bands, topology and symmetry has led to the discovery of exotic phases of matter, including fractional Chern insulators, correlated moir\'e topological superconductors, and Dirac and Weyl semimetals. Correlated metals, such as those present in Kondo lattices, rely on the screening of local moments by a sea of non-magnetic conduction electrons. Here, we report on a unique topological Kondo lattice compound, CeCo2P2, where the Kondo effect - whose existence under the magnetic Co phase is protected by PT symmetry - coexists with antiferromagnetic order emerging from the flat bands associated with the Co atoms. Remarkably, this is the only known Kondo lattice compound where magnetic order occurs in non-heavy electrons, and puzzlingly, at a temperature significantly higher than that of the Kondo effect. Furthermore, at low temperatures, the emergence of the Kondo effect, in conjunction with a glide-mirror-z symmetry, results in a nodal line protected by bulk topology near the Fermi energy. These unusual properties, arising from the interplay between itinerant and correlated electrons from different constituent elements, lead to novel quantum phases beyond the celebrated topological Kondo insulators and Weyl Kondo semimetals. CeCo2P2 thus provides an ideal platform for investigating narrow bands, topology, magnetism, and the Kondo effect in strongly correlated electron systems., Comment: 17pages,4 figures

AI systems crucially rely on human ratings, but these ratings are often aggregated, obscuring the inherent diversity of perspectives in real-world phenomenon. This is particularly concerning when evaluating the safety of generative AI, where perceptions and associated harms can vary significantly across socio-cultural contexts. While recent research has studied the impact of demographic differences on annotating text, there is limited understanding of how these subjective variations affect multimodal safety in generative AI. To address this, we conduct a large-scale study employing highly-parallel safety ratings of about 1000 text-to-image (T2I) generations from a demographically diverse rater pool of 630 raters balanced across 30 intersectional groups across age, gender, and ethnicity. Our study shows that (1) there are significant differences across demographic groups (including intersectional groups) on how severe they assess the harm to be, and that these differences vary across different types of safety violations, (2) the diverse rater pool captures annotation patterns that are substantially different from expert raters trained on specific set of safety policies, and (3) the differences we observe in T2I safety are distinct from previously documented group level differences in text-based safety tasks. To further understand these varying perspectives, we conduct a qualitative analysis of the open-ended explanations provided by raters. This analysis reveals core differences into the reasons why different groups perceive harms in T2I generations. Our findings underscore the critical need for incorporating diverse perspectives into safety evaluation of generative AI ensuring these systems are truly inclusive and reflect the values of all users., Comment: 20 pages, 7 figures

The electrical response of ferroelectric domain walls is often influenced by their geometry underneath the sample surface. Tomographic imaging in these material systems has therefore become increasingly important for its ability to correlate the surface-level functional response with subsurface domain microstructure. In this context, AFM-based tomography emerges as a compelling choice because of its simplicity, high resolution and robust contrast mechanism. However, to date, the technique has been implemented in a limited number of ferroelectric materials, typically to depths of a few hundred nanometers or on relatively soft materials, resulting in an unclear understanding of its capabilities and limitations. In this work, AFM tomography is carried out in YbMnO3, mapping its complex domain microstructure up to a depth of around 1.8 um along with its current pathways. A model is presented, describing the impact of interconnected domain walls within the network, which act as current dividers and codetermine how currents distribute. Finally, challenges such as tip-blunting and subsurface amorphisation are identified through TEM studies, and strategies to address them are also put forward. This study highlights the potential of AFM tomography and could spur interest within the ferroics community for its use in the investigation of similar material systems.

Let $S$ be an independent set of a simple undirected graph $G$. Suppose that each vertex of $S$ has a token placed on it. The tokens are allowed to be moved, one at a time, by sliding along the edges of $G$, so that after each move, the vertices having tokens always form an independent set of $G$. We would like to determine whether the tokens can be eventually brought to stay on the vertices of another independent set $S'$ of $G$ in this manner. In other words, we would like to decide if we can transform $S$ into $S'$ through a sequence of steps, each of which involves substituting a vertex in the current independent set with one of its neighbours to obtain another independent set. This problem of determining if one independent set of a graph ``is reachable'' from another independent set of it is known to be PSPACE-hard even for split graphs, planar graphs, and graphs of bounded treewidth. Polynomial time algorithms have been obtained for certain graph classes like trees, interval graphs, claw-free graphs, and bipartite permutation graphs. We present a polynomial time algorithm for the problem on block graphs, which are the graphs in which every maximal 2-connected subgraph is a clique. Our algorithm is the first generalization of the known polynomial time algorithm for trees to a larger class of graphs (note that trees form a proper subclass of block graphs)., Comment: 36 pages, 5 figures

We introduce a novel representation named as the unified gripper coordinate space for grasp synthesis of multiple grippers. The space is a 2D surface of a sphere in 3D using longitude and latitude as its coordinates, and it is shared for all robotic grippers. We propose a new algorithm to map the palm surface of a gripper into the unified gripper coordinate space, and design a conditional variational autoencoder to predict the unified gripper coordinates given an input object. The predicted unified gripper coordinates establish correspondences between the gripper and the object, which can be used in an optimization problem to solve the grasp pose and the finger joints for grasp synthesis. We demonstrate that using the unified gripper coordinate space improves the success rate and diversity in the grasp synthesis of multiple grippers., Comment: 7 pages, 8 figures, 2 tables. Project page available at https://irvlutd.github.io/RobotFingerPrint

Orthorhombic pseudobrookites CoTi$_2$O$_5$ and FeTi$_2$O$_5$ have a low-symmetry crystal structure comprising magnetic Co$^{2+}$/Fe$^{2+}$ ions and nonmagnetic Ti$^{4+}$ ions, where the orbital-nondegenerate Co$^{2+}$/Fe$^{2+}$ ions form one-dimensional chains running along the orthorhombic $a$ axis. These compounds undergo an antiferromagnetic phase transition at $T_N \sim$ 26 K for CoTi$_2$O$_5$ and $T_N \sim$ 40 K for FeTi$_2$O$_5$. We perform ultrasound velocity measurements on single crystals of CoTi$_2$O$_5$ and FeTi$_2$O$_5$. The measurements of these compounds reveal that the symmetry-lowering elastic modes of shear elastic moduli exhibit unusual elastic softness in the paramagnetic phase above $T_N$. This elastic softness indicates the presence of spin-lattice-coupled fluctuations above $T_N$ that should be a precursor to the symmetry-lowering lattice distortion at $T_N$. Furthermore, it is revealed that the magnitude of the unusual elastic softness is larger in CoTi$_2$O$_5$ than in FeTi$_2$O$_5$, which indicates that the spin-lattice coupling is stronger in CoTi$_2$O$_5$ than in FeTi$_2$O$_5$. The present study suggests that CoTi$_2$O$_5$ and FeTi$_2$O$_5$ are unique spin Jahn--Teller systems with low crystal symmetry, where, although the nature of exchange interactions is quasi-one-dimensional, the three-dimensional spin-lattice coupling releases the frustration by further lowering the crystal symmetry., Comment: 9 pages, 7 figures

Anderson's groundbreaking ideas of resonant valence bond (RVB) liquid and spin-charge separation initiated a transformative shift in modern physics. Extensive implications influenced a broad spectrum of fields, from high-temperature superconductors to quantum computing, and gave birth to key concepts in physics, such as quantum spin liquids, emergent gauge symmetries, topological order, and fractionalisation. Despite extensive efforts to demonstrate the existence of an RVB phase in the Hubbard model, a definitive realisation has proven elusive. Here, we present a concise, realistic, and elegant solution to this longstanding problem by demonstrating analytically that an RVB spin liquid, exhibiting spin-charge separation, emerges as the ground state of doped Mott insulators on corner-sharing tetrahedral lattices with frustrated hopping near half-filling -- a manifestation of the counter-Nagaoka effect. We confirm numerically that this result holds for finite-size systems, finite dopant density, and small exchange interactions. While much attention has been devoted to the emergence of new states from geometrically frustrated interactions, our work demonstrates that kinetic energy frustration in doped Mott insulators may be pivotal to stabilise robust, topologically ordered states in real materials.

Deep neural networks have demonstrated exceptional efficacy in stroke lesion segmentation. However, the delineation of small lesions, critical for stroke diagnosis, remains a challenge. In this study, we propose two straightforward yet powerful approaches that can be seamlessly integrated into a variety of networks: Multi-Size Labeling (MSL) and Distance-Based Labeling (DBL), with the aim of enhancing the segmentation accuracy of small lesions. MSL divides lesion masks into various categories based on lesion volume while DBL emphasizes the lesion boundaries. Experimental evaluations on the Anatomical Tracings of Lesions After Stroke (ATLAS) v2.0 dataset showcase that an ensemble of MSL and DBL achieves consistently better or equal performance on recall (3.6% and 3.7%), F1 (2.4% and 1.5%), and Dice scores (1.3% and 0.0%) compared to the top-1 winner of the 2022 MICCAI ATLAS Challenge on both the subset only containing small lesions and the entire dataset, respectively. Notably, on the mini-lesion subset, a single MSL model surpasses the previous best ensemble strategy, with enhancements of 1.0% and 0.3% on F1 and Dice scores, respectively. Our code is available at: https://github.com/nadluru/StrokeLesSeg.

Chirality is a pervasive form of symmetry that is intimately connected to the physical properties of solids, as well as the chemical and biological activity of molecular systems. However, its control with light is challenging, because inducing chirality in a non-chiral material requires that all mirrors and all roto-inversions be simultaneously broken. Electromagnetic fields exert only oscillatory forces that vanish on average, mostly leading to entropy increase that does not break symmetries, per se. Here, we show that chirality of either handedness can be generated in the non-chiral piezoelectric material BPO$_4$, in which two compensated sub-structures of opposite handedness coexist within the same unit cell. By resonantly driving either one of two orthogonal, doubly degenerate vibrational modes at Terahertz frequency, we rectify the lattice distortion and exert a displacive force onto the crystal. The staggered chirality is in this way uncompensated in either direction, inducing chiral structure with either handedness. The rotary power of the photo-induced phases is comparable to the static value of prototypical chiral alpha-quartz, limited by the strength of the pump laser pulse., Comment: 27 pages, including Supplementary Materials

Text-to-Image (T2I) models are being increasingly adopted in diverse global communities where they create visual representations of their unique cultures. Current T2I benchmarks primarily focus on faithfulness, aesthetics, and realism of generated images, overlooking the critical dimension of cultural competence. In this work, we introduce a framework to evaluate cultural competence of T2I models along two crucial dimensions: cultural awareness and cultural diversity, and present a scalable approach using a combination of structured knowledge bases and large language models to build a large dataset of cultural artifacts to enable this evaluation. In particular, we apply this approach to build CUBE (CUltural BEnchmark for Text-to-Image models), a first-of-its-kind benchmark to evaluate cultural competence of T2I models. CUBE covers cultural artifacts associated with 8 countries across different geo-cultural regions and along 3 concepts: cuisine, landmarks, and art. CUBE consists of 1) CUBE-1K, a set of high-quality prompts that enable the evaluation of cultural awareness, and 2) CUBE-CSpace, a larger dataset of cultural artifacts that serves as grounding to evaluate cultural diversity. We also introduce cultural diversity as a novel T2I evaluation component, leveraging quality-weighted Vendi score. Our evaluations reveal significant gaps in the cultural awareness of existing models across countries and provide valuable insights into the cultural diversity of T2I outputs for under-specified prompts. Our methodology is extendable to other cultural regions and concepts, and can facilitate the development of T2I models that better cater to the global population., Comment: To appear at NeurIPS 2024 (D&B)

Understanding how light modifies long-range order is key to improve our ability to control material functionality on an ultrafast timescale. Transient spatial heterogeneity has been proposed in many materials, but isolating the dynamics of different regions experimentally has been challenging. Here we address this issue and measure the dynamics of orbital order melting in the layered manganite, La0.5Sr1.5MnO4, and isolate the surface dynamics from the bulk for the first time. Bulk measurements show orbital order is rapidly suppressed, but the correlation length surprisingly increases. However, the surface dynamics, show a stronger suppression and a significant decrease in correlation length. By isolating the surface changes, we find that light preferentially melts a less ordered surface and the loss of long-range order is likely driven by the formation of local and disordered polarons. Melting the disordered surface effectively increases the average correlation of the bulk probed volume, resolving the contradictory response. These results show that surface scattering methods are necessary to understand both surface and bulk dynamics in heterogeneous materials., Comment: 22 pages, 8 figures

Segregation encodes information about society, such as social cohesion, mixing, and inequality. However, most past and current studies tackled socioeconomic (SE) segregation by analyzing static aggregated mobility networks, often without considering further individual features beyond income and, most importantly, without distinguishing individual-level from location-based income. Accessing individual-level income may help mapping macroscopic behavior into more granular mobility patterns, hence impacting segregation estimates. Here we combine a mobile phone dataset of daily mobility flows across Spanish districts stratified and adjusted by age, gender and income with census data of districts median income. We build mobility-based SE assortativity matrices for multiple demographics and observe mobility patterns of three income groups with respect to location-based SE classes. We find that SE assortativity differs when isolating the mobility of specific income groups: we observe that groups prefer to visit areas with higher average income than their own, which we call preferential mobility. Our analysis suggests substantial differences between weekdays and weekends SE assortativity by age class, with weekends characterized by higher SE assortativity. Our modeling approach shows that the radiation model, which typically performs best at reproducing inter-municipal population mobility, best fits middle income and middle-aged flows, while performing worse on young and low income groups. Our double-sided approach, focusing on assortativity patterns and mobility modeling, suggests that state of the art mobility models fail at capturing preferential mobility behavior. Overall, our work indicates that mobility models considering the interplay of SE preferential behavior, age and gender gaps may sensibly improve the state of the art models performance., Comment: 11 pages, 3 figures, complexity72h

Algorithmic fairness has emerged as a critical concern in artificial intelligence (AI) research. However, the development of fair AI systems is not an objective process. Fairness is an inherently subjective concept, shaped by the values, experiences, and identities of those involved in research and development. To better understand the norms and values embedded in current fairness research, we conduct a meta-analysis of algorithmic fairness papers from two leading conferences on AI fairness and ethics, AIES and FAccT, covering a final sample of 139 papers over the period from 2018 to 2022. Our investigation reveals two concerning trends: first, a US-centric perspective dominates throughout fairness research; and second, fairness studies exhibit a widespread reliance on binary codifications of human identity (e.g., "Black/White", "male/female"). These findings highlight how current research often overlooks the complexities of identity and lived experiences, ultimately failing to represent diverse global contexts when defining algorithmic bias and fairness. We discuss the limitations of these research design choices and offer recommendations for fostering more inclusive and representative approaches to fairness in AI systems, urging a paradigm shift that embraces nuanced, global understandings of human identity and values.

We introduce a novel, data-driven topological data analysis (TDA) approach for embedding brain networks into a lower-dimensional space in quantifying the dynamics of temporal lobe epilepsy (TLE) obtained from resting-state functional magnetic resonance imaging (rs-fMRI). This embedding facilitates the orthogonal projection of 0D and 1D topological features, allowing for the visualization and modeling of the dynamics of functional human brain networks in a resting state. We then quantify the topological disparities between networks to determine the coordinates for embedding. This framework enables us to conduct a coherent statistical inference within the embedded space. Our results indicate that brain network topology in TLE patients exhibits increased rigidity in 0D topology but more rapid flections compared to that of normal controls in 1D topology.

We study the problem of synthesising a two-user broadcast channel using a common message, where each output terminal shares an independent source of randomness with the input terminal. This generalises two problems studied in the literature (Cuff, IEEE Trans. Inform. Theory, 2013; Kurri et.al., IEEE Trans. Inform. Theory, 2021). We give an inner bound on the tradeoff region between the rates of communication and shared randomness, and a lower bound on the minimum communication rate. Although the bounds presented here are not tight in general, they are tight for some special cases, including the aforementioned problems.

We introduce the study of information leakage through \emph{guesswork}, the minimum expected number of guesses required to guess a random variable. In particular, we define \emph{maximal guesswork leakage} as the multiplicative decrease, upon observing $Y$, of the guesswork of a randomized function of $X$, maximized over all such randomized functions. We also study a pointwise form of the leakage which captures the leakage due to the release of a single realization of $Y$. We also study these two notions of leakage with oblivious (or memoryless) guessing. We obtain closed-form expressions for all these leakage measures, with the exception of one. Specifically, we are able to obtain closed-form expression for maximal guesswork leakage for the binary erasure source only; deriving expressions for arbitrary sources appears challenging. Some of the consequences of our results are -- a connection between guesswork and differential privacy and a new operational interpretation to maximal $\alpha$-leakage in terms of guesswork., Comment: 6 pages. Extended version of a paper accepted to ISIT 2024

While human annotations play a crucial role in language technologies, annotator subjectivity has long been overlooked in data collection. Recent studies that have critically examined this issue are often situated in the Western context, and solely document differences across age, gender, or racial groups. As a result, NLP research on subjectivity have overlooked the fact that individuals within demographic groups may hold diverse values, which can influence their perceptions beyond their group norms. To effectively incorporate these considerations into NLP pipelines, we need datasets with extensive parallel annotations from various social and cultural groups. In this paper we introduce the \dataset dataset: a large-scale cross-cultural dataset of parallel annotations for offensive language in over 4.5K sentences annotated by a pool of over 4k annotators, balanced across gender and age, from across 21 countries, representing eight geo-cultural regions. The dataset contains annotators' moral values captured along six moral foundations: care, equality, proportionality, authority, loyalty, and purity. Our analyses reveal substantial regional variations in annotators' perceptions that are shaped by individual moral values, offering crucial insights for building pluralistic, culturally sensitive NLP models.

Generative language models are transforming our digital ecosystem, but they often inherit societal biases, for instance stereotypes associating certain attributes with specific identity groups. While whether and how these biases are mitigated may depend on the specific use cases, being able to effectively detect instances of stereotype perpetuation is a crucial first step. Current methods to assess presence of stereotypes in generated language rely on simple template or co-occurrence based measures, without accounting for the variety of sentential contexts they manifest in. We argue that understanding the sentential context is crucial for detecting instances of generalization. We distinguish two types of generalizations: (1) language that merely mentions the presence of a generalization ("people think the French are very rude"), and (2) language that reinforces such a generalization ("as French they must be rude"), from non-generalizing context ("My French friends think I am rude"). For meaningful stereotype evaluations, we need to reliably distinguish such instances of generalizations. We introduce the new task of detecting generalization in language, and build GeniL, a multilingual dataset of over 50K sentences from 9 languages (English, Arabic, Bengali, Spanish, French, Hindi, Indonesian, Malay, and Portuguese) annotated for instances of generalizations. We demonstrate that the likelihood of a co-occurrence being an instance of generalization is usually low, and varies across different languages, identity groups, and attributes. We build classifiers to detect generalization in language with an overall PR-AUC of 58.7, with varying degrees of performance across languages. Our research provides data and tools to enable a nuanced understanding of stereotype perpetuation, a crucial step towards more inclusive and responsible language technologies.

We report a resonant inelastic X-ray scattering study on a single crystal of a non-stoichiometric pyrochlore iridate Tb$_{2+x}$Ir$_{2-x}$O$_{7-y}$ ($x \simeq 0.25$) that magnetically orders at $T_{\rm{N}}\simeq 50$ K. We find that the strength of the spin-orbit coupling and the trigonal distortion of the IrO$_6$ octahedra are comparable with the ones obtained in other pyrochlore iridates. We observe a propagating gapped magnon mode at low energy, and model it using a Hamiltonian consisting of a Heisenberg exchange [$J = 16.2(9)$ meV] and Dzyaloshinskii-Moriya interactions [$D = 5.2(3)$ meV], which shows the robustness of interactions despite Tb-stuffing at the Ir-site. Strikingly, the ratio $D/J = 0.32(3)$ supports possible non-trivial topological magnon band crossing. This material may thus host coexisting fermionic and bosonic topology, with potential for manipulating electronic and magnonic topological bands thanks to the $d-f$ interaction., Comment: manuscript: 7 pages, 4 figures; Supplemental Material: 6 pages, 7 figures, 1 table. Accepted at PRB Letters (2024)

Magnetic resonance imaging (MRI) is the standard modality to understand human brain structure and function in vivo (antemortem). Decades of research in human neuroimaging has led to the widespread development of methods and tools to provide automated volume-based segmentations and surface-based parcellations which help localize brain functions to specialized anatomical regions. Recently ex vivo (postmortem) imaging of the brain has opened-up avenues to study brain structure at sub-millimeter ultra high-resolution revealing details not possible to observe with in vivo MRI. Unfortunately, there has been limited methodological development in ex vivo MRI primarily due to lack of datasets and limited centers with such imaging resources. Therefore, in this work, we present one-of-its-kind dataset of 82 ex vivo T2w whole brain hemispheres MRI at 0.3 mm isotropic resolution spanning Alzheimer's disease and related dementias. We adapted and developed a fast and easy-to-use automated surface-based pipeline to parcellate, for the first time, ultra high-resolution ex vivo brain tissue at the native subject space resolution using the Desikan-Killiany-Tourville (DKT) brain atlas. This allows us to perform vertex-wise analysis in the template space and thereby link morphometry measures with pathology measurements derived from histology. We will open-source our dataset docker container, Jupyter notebooks for ready-to-use out-of-the-box set of tools and command line options to advance ex vivo MRI clinical brain imaging research on the project webpage.

We introduce a large-scale dataset named MultiGripperGrasp for robotic grasping. Our dataset contains 30.4M grasps from 11 grippers for 345 objects. These grippers range from two-finger grippers to five-finger grippers, including a human hand. All grasps in the dataset are verified in the robot simulator Isaac Sim to classify them as successful and unsuccessful grasps. Additionally, the object fall-off time for each grasp is recorded as a grasp quality measurement. Furthermore, the grippers in our dataset are aligned according to the orientation and position of their palms, allowing us to transfer grasps from one gripper to another. The grasp transfer significantly increases the number of successful grasps for each gripper in the dataset. Our dataset is useful to study generalized grasp planning and grasp transfer across different grippers. Data, code and videos for the project are available at https://irvlutd.github.io/MultiGripperGrasp, Comment: Published in IROS 2024

We extend previous inelastic neutron scattering results on the geometrically frustrated antiferromagnet hexagonal-YMnO$_3$, which has been suggested to belong to the class of classical spin liquids. We extend the energy transfer coverage of the diffuse signal up to 6.9 meV within a wide temperature range around the ordering temperature, $T_\mathrm{N}$. The two distinct diffuse signals in the a-b plane, the signal localized at $\Gamma$' and the scattering intensity connecting $\Gamma$' points over the M', are shown to be only weakly energy dependent. In addition, an external magnetic field of up to 10.5 T applied along c is shown to have no effect on the diffuse signal. In the orthogonal scattering plane, the signals are shown to be dependent on l only through the magnetic form factor, showing that the correlations are purely two-dimensional, and supporting its origin to be the frustrated Mn$^{3+}$ triangles. This result is corroborated by atomistic spin dynamics simulations showing similar scattering vector and temperature behaviours. Lastly, data for the spin wave scattering in the (h, 0, l) plane allow for a discussion of the magnetic ground state where better agreement is found between the data and an ordered structure of the $\Gamma_1$ or $\Gamma_3$ symmetry, albeit crystal electric field arguments dismisses the $\Gamma_1$ as possibility.

While generative multilingual models are rapidly being deployed, their safety and fairness evaluations are largely limited to resources collected in English. This is especially problematic for evaluations targeting inherently socio-cultural phenomena such as stereotyping, where it is important to build multi-lingual resources that reflect the stereotypes prevalent in respective language communities. However, gathering these resources, at scale, in varied languages and regions pose a significant challenge as it requires broad socio-cultural knowledge and can also be prohibitively expensive. To overcome this critical gap, we employ a recently introduced approach that couples LLM generations for scale with culturally situated validations for reliability, and build SeeGULL Multilingual, a global-scale multilingual dataset of social stereotypes, containing over 25K stereotypes, spanning 20 languages, with human annotations across 23 regions, and demonstrate its utility in identifying gaps in model evaluations. Content warning: Stereotypes shared in this paper can be offensive.

Ocular diseases, including diabetic retinopathy and glaucoma, present a significant public health challenge due to their high prevalence and potential for causing vision impairment. Early and accurate diagnosis is crucial for effective treatment and management. In recent years, deep learning models have emerged as powerful tools for analysing medical images, such as retina imaging. However, challenges persist in model relibability and uncertainty estimation, which are critical for clinical decision-making. This study leverages the probabilistic framework of Generative Flow Networks (GFlowNets) to learn the posterior distribution over latent discrete dropout masks for the classification and analysis of ocular diseases using fundus images. We develop a robust and generalizable method that utilizes GFlowOut integrated with ResNet18 and ViT models as the backbone in identifying various ocular conditions. This study employs a unique set of dropout masks - none, random, bottomup, and topdown - to enhance model performance in analyzing these fundus images. Our results demonstrate that our learnable probablistic latents significantly improves accuracy, outperforming the traditional dropout approach. We utilize a gradient map calculation method, Grad-CAM, to assess model explainability, observing that the model accurately focuses on critical image regions for predictions. The integration of GFlowOut in neural networks presents a promising advancement in the automated diagnosis of ocular diseases, with implications for improving clinical workflows and patient outcomes.

Recent studies have shown that Text-to-Image (T2I) model generations can reflect social stereotypes present in the real world. However, existing approaches for evaluating stereotypes have a noticeable lack of coverage of global identity groups and their associated stereotypes. To address this gap, we introduce the ViSAGe (Visual Stereotypes Around the Globe) dataset to enable the evaluation of known nationality-based stereotypes in T2I models, across 135 nationalities. We enrich an existing textual stereotype resource by distinguishing between stereotypical associations that are more likely to have visual depictions, such as `sombrero', from those that are less visually concrete, such as 'attractive'. We demonstrate ViSAGe's utility through a multi-faceted evaluation of T2I generations. First, we show that stereotypical attributes in ViSAGe are thrice as likely to be present in generated images of corresponding identities as compared to other attributes, and that the offensiveness of these depictions is especially higher for identities from Africa, South America, and South East Asia. Second, we assess the stereotypical pull of visual depictions of identity groups, which reveals how the 'default' representations of all identity groups in ViSAGe have a pull towards stereotypical depictions, and that this pull is even more prominent for identity groups from the Global South. CONTENT WARNING: Some examples contain offensive stereotypes., Comment: Association for Computational Linguistics (ACL) 2024