Search

A diffusive system coupled to unequal boundary reservoirs eventually reaches a non-equilibrium steady state. While the full-counting-statistics of current in these non-equilibrium states are well understood for generic systems, results for steady-state density fluctuations remain limited to a few integrable models. By employing an exact solution of the Macroscopic Fluctuation Theory, we characterize steady-state density fluctuations in terms of large deviations for a class of systems. Additionally, we quantitatively describe the development of these rare fluctuations. For generic diffusive systems in arbitrary dimensions, we use a perturbation around the equilibrium state to solve for the large deviation functional and the corresponding path to fluctuations, up to non-trivial orders where the non-locality of fluctuations unveils., Comment: 6 pages, 2 figures, 1 table (+ 5 supplementary pages)

We present a class of stochastic processes in which the large deviation functions of time-integrated observables exhibit singularities that relate to dynamical phase transitions of trajectories. These illustrative examples include Brownian motion with a death rate or in the presence of an absorbing wall, for which we consider a set of empirical observables such as the net displacement, local time, residence time, and area under the trajectory. Using a backward Fokker-Planck approach, we derive the large deviation functions of these observables, and demonstrate how singularities emerge from a competition between survival and diffusion. Furthermore, we analyse this scenario using an alternative approach with tilted operators, showing that at the singular point, the effective dynamics undergoes an abrupt transition. Extending this approach, we show that similar transitions may generically arise in Markov chains with transient states. This scenario is robust and generalizable for non-Markovian dynamics and for many-body systems, potentially leading to multiple dynamical phase transitions., Comment: 21 pages, 21 figures. For a supplemental Mathematica notebook (Ref[84]), see https://drive.google.com/file/d/1cAcXngt6wE7GCo0u-ltjShSmXtswgKPO/view?usp=sharing

Theory of Mind (ToM) refers to the cognitive ability to infer and attribute mental states to oneself and others. As large language models (LLMs) are increasingly evaluated for social and cognitive capabilities, it remains unclear to what extent these models demonstrate ToM across diverse languages and cultural contexts. In this paper, we introduce a comprehensive study of multilingual ToM capabilities aimed at addressing this gap. Our approach includes two key components: (1) We translate existing ToM datasets into multiple languages, effectively creating a multilingual ToM dataset and (2) We enrich these translations with culturally specific elements to reflect the social and cognitive scenarios relevant to diverse populations. We conduct extensive evaluations of six state-of-the-art LLMs to measure their ToM performance across both the translated and culturally adapted datasets. The results highlight the influence of linguistic and cultural diversity on the models' ability to exhibit ToM, and questions their social reasoning capabilities. This work lays the groundwork for future research into enhancing LLMs' cross-cultural social cognition and contributes to the development of more culturally aware and socially intelligent AI systems. All our data and code are publicly available.

Hepatitis B virus (HBV) is considered as etiological agent of the lethal liver disease hepatitis B. Globally, hepatitis B is recognized as one of the prevailing infectious diseases with a significant impact on human health. In spite of being non-infectious in nature, sub-viral particles (SVPs) , composed with mainly viral surface proteins, play critical roles in the persistence and progression of the infection. Although the understanding on the functions of these non-infectious SVPs remains limited and incomplete. In this study, a mathematical model is proposed for the first time by incorporating the roles of SVPs and including the effects of capsids recycling. The impacts of spatial mobility of capsids, viruses, SVPs and antibodies are also taken into account in this model. Overall, this model carry unique characteristics in the context of this viral infection. This study investigates the changes in the dynamics of infection considering both single-point as well as multi-point infection initial condition. As a result, it is observed that SVPs can significantly enhance intracellular viral replication and gene expression by reducing the neutralization of virus particles by antibodies. The recycling of capsids substantially increase the concentration of SVPs. The experiments which are carried out on single-point as well as multi-point infection initial conditions show that if the liver is infected at more than one point, infection propagates rapidly. The outcomes of the proposed model strongly recommend that it is imperative to consider the diffusion term while HBV infection dynamics are illustrated.

Entropic optimal transport offers a computationally tractable approximation to the classical problem. In this note, we study the approximation rate of the entropic optimal transport map (in approaching the Brenier map) when the regularization parameter $\varepsilon$ tends to zero in the semidiscrete setting, where the input measure is absolutely continuous while the output is finitely discrete. Previous work shows that the approximation rate is $O(\sqrt{\varepsilon})$ under the $L^2$-norm with respect to the input measure. In this work, we establish faster, $O(\varepsilon^2)$ rates up to polylogarithmic factors, under the dual Lipschitz norm, which is weaker than the $L^2$-norm. For the said dual norm, the $O(\varepsilon^2)$ rate is sharp. As a corollary, we derive a central limit theorem for the entropic estimator for the Brenier map in the dual Lipschitz space when the regularization parameter tends to zero as the sample size increases.

Recent advancements in large language models have demonstrated their potential in numerous medical applications, particularly in automating clinical trial matching for translational research and enhancing medical question answering for clinical decision support. However, our study shows that incorporating non decisive sociodemographic factors such as race, sex, income level, LGBT+ status, homelessness, illiteracy, disability, and unemployment into the input of LLMs can lead to incorrect and harmful outputs for these populations. These discrepancies risk exacerbating existing health disparities if LLMs are widely adopted in healthcare. To address this issue, we introduce EquityGuard, a novel framework designed to detect and mitigate the risk of health inequities in LLM based medical applications. Our evaluation demonstrates its efficacy in promoting equitable outcomes across diverse populations.

Active-matter systems are inherently out-of-equilibrium and perform mechanical work by utilizing their internal energy sources. Breakdown of time-reversal symmetry (BTRS) is a hallmark of such dissipative non-equilibrium dynamics. We introduce a robust, experimentally accessible, noninvasive, quantitative measure of BTRS in terms of the Kullback-Leibler divergence in collision events, demonstrated in our novel artificial active matter, comprised of battery-powered spherical rolling robots whose energetics in different modes of motion can be measured with high precision. Our dimensionless measure characterizes how dissipation and internal energetics are influenced by kinematic constraints from interactions with the environment. We propose this measure of BTRS as an empirical estimate of the distance from equilibrium. An energetic insight into this departure of active matter from equilibrium comes from our demonstration of a non-trivial fluctuation symmetry, which reveals a potentially universal thermodynamic characteristic of active energetics. As a many-body consequence of BTRS in our experimental active system, we demonstrate the emergence of activity-induced herding, which has no equilibrium analogue., Comment: 9 pages, 7 figures, supplemental information attached

In this article, we revisit Weibel's conjecture for twisted $K$-theory. We also examine the vanishing of twisted negative $K$-groups for Pr\"{u}fer domains. Furthermore, we observe that the homotopy invariance of twisted $K$-theory holds for (finite-dimensional) Pr\"{u}fer domains.

Due to emergent capabilities, large language models (LLMs) have been utilized as language-based agents to perform a variety of tasks and make decisions with an increasing degree of autonomy. These autonomous agents can understand high-level instructions, interact with their environments, and execute complex tasks using a selection of tools available to them. As the capabilities of the agents expand, ensuring their safety and trustworthiness becomes more imperative. In this study, we introduce the Athena framework which leverages the concept of verbal contrastive learning where past safe and unsafe trajectories are used as in-context (contrastive) examples to guide the agent towards safety while fulfilling a given task. The framework also incorporates a critiquing mechanism to guide the agent to prevent risky actions at every step. Furthermore, due to the lack of existing benchmarks on the safety reasoning ability of LLM-based agents, we curate a set of 80 toolkits across 8 categories with 180 scenarios to provide a safety evaluation benchmark. Our experimental evaluation, with both closed- and open-source LLMs, indicates verbal contrastive learning and interaction-level critiquing improve the safety rate significantly., Comment: 9 pages, 2 figures, 4 tables

Large language models (LLMs) have brought autonomous agents closer to artificial general intelligence (AGI) due to their promising generalization and emergent capabilities. There is, however, a lack of studies on how LLM-based agents behave, why they could potentially fail, and how to improve them, particularly in demanding real-world planning tasks. In this paper, as an effort to fill the gap, we present our study using a realistic benchmark, TravelPlanner, where an agent must meet multiple constraints to generate accurate plans. We leverage this benchmark to address four key research questions: (1) are LLM agents robust enough to lengthy and noisy contexts when it comes to reasoning and planning? (2) can few-shot prompting adversely impact the performance of LLM agents in scenarios with long context? (3) can we rely on refinement to improve plans, and (4) can fine-tuning LLMs with both positive and negative feedback lead to further improvement? Our comprehensive experiments indicate that, firstly, LLMs often fail to attend to crucial parts of a long context, despite their ability to handle extensive reference information and few-shot examples; secondly, they still struggle with analyzing the long plans and cannot provide accurate feedback for refinement; thirdly, we propose Feedback-Aware Fine-Tuning (FAFT), which leverages both positive and negative feedback, resulting in substantial gains over Supervised Fine-Tuning (SFT). Our findings offer in-depth insights to the community on various aspects related to real-world planning applications., Comment: 13 pages, 2 figures, 4 tables

Recent advancements in deep reinforcement learning (DRL) techniques have sparked its multifaceted applications in the automation sector. Managing complex decision-making problems with DRL encourages its use in the nuclear industry for tasks such as optimizing radiation exposure to the personnel during normal operating conditions and potential accidental scenarios. However, the lack of efficient reward function and effective exploration strategy thwarted its implementation in the development of radiation-aware autonomous unmanned aerial vehicle (UAV) for achieving maximum radiation protection. Here, in this article, we address these intriguing issues and introduce a deep Q-learning based architecture (RadDQN) that operates on a radiation-aware reward function to provide time-efficient minimum radiation-exposure pathway in a radiation zone. We propose a set of unique exploration strategies that fine-tune the extent of exploration and exploitation based on the state-wise variation in radiation exposure during training. Further, we benchmark the predicted path with grid-based deterministic method. We demonstrate that the formulated reward function in conjugation with adequate exploration strategy is effective in handling several scenarios with drastically different radiation field distributions. When compared to vanilla DQN, our model achieves a superior convergence rate and higher training stability., Comment: 12 pages, 7 main figures, code link (GitHub)

The influence of Large Language Models (LLMs) is rapidly growing, automating more jobs over time. Assessing the fairness of LLMs is crucial due to their expanding impact. Studies reveal the reflection of societal norms and biases in LLMs, which creates a risk of propagating societal stereotypes in downstream tasks. Many studies on bias in LLMs focus on gender bias in various NLP applications. However, there's a gap in research on bias in emotional attributes, despite the close societal link between emotion and gender. This gap is even larger for low-resource languages like Bangla. Historically, women are associated with emotions like empathy, fear, and guilt, while men are linked to anger, bravado, and authority. This pattern reflects societal norms in Bangla-speaking regions. We offer the first thorough investigation of gendered emotion attribution in Bangla for both closed and open source LLMs in this work. Our aim is to elucidate the intricate societal relationship between gender and emotion specifically within the context of Bangla. We have been successful in showing the existence of gender bias in the context of emotions in Bangla through analytical methods and also show how emotion attribution changes on the basis of gendered role selection in LLMs. All of our resources including code and data are made publicly available to support future research on Bangla NLP. Warning: This paper contains explicit stereotypical statements that many may find offensive., Comment: Accepted at the 5th Workshop on Gender Bias in Natural Language Processing at the ACL 2024 Conference

The rapid growth of Large Language Models (LLMs) has put forward the study of biases as a crucial field. It is important to assess the influence of different types of biases embedded in LLMs to ensure fair use in sensitive fields. Although there have been extensive works on bias assessment in English, such efforts are rare and scarce for a major language like Bangla. In this work, we examine two types of social biases in LLM generated outputs for Bangla language. Our main contributions in this work are: (1) bias studies on two different social biases for Bangla, (2) a curated dataset for bias measurement benchmarking and (3) testing two different probing techniques for bias detection in the context of Bangla. This is the first work of such kind involving bias assessment of LLMs for Bangla to the best of our knowledge. All our code and resources are publicly available for the progress of bias related research in Bangla NLP., Comment: Accepted at The First Workshop on Language Models for Low-Resource Languages (LoResLM) at COLING 2025

Pretrained language models inherently exhibit various social biases, prompting a crucial examination of their social impact across various linguistic contexts due to their widespread usage. Previous studies have provided numerous methods for intrinsic bias measurements, predominantly focused on high-resource languages. In this work, we aim to extend these investigations to Bangla, a low-resource language. Specifically, in this study, we (1) create a dataset for intrinsic gender bias measurement in Bangla, (2) discuss necessary adaptations to apply existing bias measurement methods for Bangla, and (3) examine the impact of context length variation on bias measurement, a factor that has been overlooked in previous studies. Through our experiments, we demonstrate a clear dependency of bias metrics on context length, highlighting the need for nuanced considerations in Bangla bias analysis. We consider our work as a stepping stone for bias measurement in the Bangla Language and make all of our resources publicly available to support future research., Comment: Accepted in Findings of ACL, 2024

Quantum architecture Search (QAS) is a promising direction for optimization and automated design of quantum circuits towards quantum advantage. Recent techniques in QAS emphasize Multi-Layer Perceptron (MLP)-based deep Q-networks. However, their interpretability remains challenging due to the large number of learnable parameters and the complexities involved in selecting appropriate activation functions. In this work, to overcome these challenges, we utilize the Kolmogorov-Arnold Network (KAN) in the QAS algorithm, analyzing their efficiency in the task of quantum state preparation and quantum chemistry. In quantum state preparation, our results show that in a noiseless scenario, the probability of success is 2 to 5 times higher than MLPs. In noisy environments, KAN outperforms MLPs in fidelity when approximating these states, showcasing its robustness against noise. In tackling quantum chemistry problems, we enhance the recently proposed QAS algorithm by integrating curriculum reinforcement learning with a KAN structure. This facilitates a more efficient design of parameterized quantum circuits by reducing the number of required 2-qubit gates and circuit depth. Further investigation reveals that KAN requires a significantly smaller number of learnable parameters compared to MLPs; however, the average time of executing each episode for KAN is higher., Comment: Code available at: https://github.com/Aqasch/KANQAS_code

Quantum steering is a kind of bipartite quantum correlations where one party's measurement remotely alters the state of another party. In an adversarial scenario, there could be a hidden variable introducing a bias in the choice of measurement settings of the parties. However, observers without access to the hidden variable are unaware of this bias. The main focus of this work is to analyze quantum steering without assuming that the parties freely choose their measurement settings. For this, we introduce the measurement-dependent (MD-)steering scenario where the measurement settings chosen by the parties are biased by an adversary. In such a scenario, we present a class of inequalities to test for MD-steerable correlations. Further, we discuss the implications of violating such inequalities in certifying randomness from quantum extremal behaviors. We also assume that an adversary might prepare an assemblage as a mixture of MD-steerable and MD-unsteerable assemblages and provide a bound on the measurement dependence for the observed correlation to remain MD-steerable., Comment: 10 pages, 5 figures

We present a fluctuating hydrodynamic description of a non-polar active lattice gas model with excluded volume interactions that exhibits motility-induced phase separation under appropriate conditions. For quasi-one dimension and higher, stability analysis of the noiseless hydrodynamics gives quantitative bounds on the phase boundary of the motility-induced phase separation in terms of spinodal and binodal. Inclusion of the multiplicative noise in the fluctuating hydrodynamics describes the exponentially decaying two-point correlations in the stationary-state homogeneous phase. Our hydrodynamic description and theoretical predictions based on it are in excellent agreement with our Monte-Carlo simulations and pseudo-spectral iteration of the hydrodynamics equations. Our construction of hydrodynamics for this model is not suitable in strictly one-dimension with single-file constraints, and we argue that this breakdown is associated with micro-phase separation., Comment: 6 pages, 6 figures, + 16 supplemental pages

Two of the most influential exact results in classical one-dimensional diffusive transport are about current statistics for the symmetric simple exclusion process in the stationary state on a finite line coupled with two unequal reservoirs at the boundary, and in the non-stationary state on an infinite line. We present the corresponding result for the intermediate geometry of a semi-infinite line coupled with a single reservoir. This result is obtained using the fluctuating hydrodynamics approach of macroscopic fluctuation theory and confirmed by rare event simulations using a cloning algorithm. Our exact result enables us to address the corresponding problem on an infinite line in the presence of a slow bond and several related problems., Comment: 8 pages, 4 figures (Supplementary Materials: 6 pages, 2 figures); Several corrections made including new numerical results and an updated Supplement

In the field of quantum computing, variational quantum algorithms (VQAs) represent a pivotal category of quantum solutions across a broad spectrum of applications. These algorithms demonstrate significant potential for realising quantum computational advantage. A fundamental aspect of VQAs involves formulating expressive and efficient quantum circuits (namely ansatz), and automating the search of such ansatz is known as quantum architecture search (QAS). Recently, reinforcement learning (RL) techniques is utilized to automate the search for ansatzes, known as RL-QAS. This study investigates RL-QAS for crafting ansatz tailored to the variational quantum state diagonalisation problem. Our investigation includes a comprehensive analysis of various dimensions, such as the entanglement thresholds of the resultant states, the impact of initial conditions on the performance of RL-agent, the phase transition behaviour of correlation in concurrence bounds, and the discrete contributions of qubits in deducing eigenvalues through conditional entropy metrics. We leverage these insights to devise an entanglement-guided admissible ansatz in QAS to diagonalise random quantum states using optimal resources. Furthermore, the methodologies presented herein offer a generalised framework for constructing reward functions within RL-QAS applicable to variational quantum algorithms., Comment: Published version

We study the Generalized Red-Blue Annulus Cover problem for two sets of points, red ($R$) and blue ($B$), where each point $p \in R\cup B$ is associated with a positive penalty ${\cal P}(p)$. The red points have non-covering penalties, and the blue points have covering penalties. The objective is to compute a circular annulus ${\cal A}$ such that the value of the function ${\cal P}({R}^{out})$ + ${\cal P}({ B}^{in})$ is minimum, where ${R}^{out} \subseteq {R}$ is the set of red points not covered by ${\cal A}$ and ${B}^{in} \subseteq {B}$ is the set of blue points covered by $\cal A$. We also study another version of this problem, where all the red points in $R$ and the minimum number of points in $B$ are covered by the circular annulus in two dimensions. We design polynomial-time algorithms for all such circular annulus problems., Comment: 6 figures

We investigate the Baryonic Faber-Jackson Relation (BFJR), examining the correlation between baryonic mass and velocity dispersion in galaxy groups and clusters. Originally analysed in elliptical galaxies, the BFJR is derivable from the empirical Radial Acceleration Relation (RAR) and MOdified Newtonian Dynamics (MOND), both showcasing a characteristic acceleration scale $g_\mathrm{\dagger}=1.2\times10^{-10}\,\mathrm{m}\,\mathrm{s}^{-2}$. Recent interpretations within MOND suggest that galaxy group dynamics can be explained solely by baryonic mass, hinting at a BFJR with $g_{\dagger}$ in these systems. To explore this BFJR, we combined X-ray and optical measurements for six galaxy clusters and 13 groups, calculating baryonic masses by combining X-ray gas and stellar mass estimates. Simultaneously, we computed spatially resolved velocity dispersion profiles from membership data using the biweight scale in radial bins. Our results indicate that the BFJR in galaxy groups, using total velocity dispersion, aligns with MOND predictions. Conversely, galaxy clusters exhibit a parallel BFJR with a larger acceleration scale. Analysis using tail velocity dispersion in galaxy groups shows a leftward deviation from the BFJR. Additionally, stacked velocity dispersion profiles reveal two distinct types: declining and flat, based on two parallel BFJRs. The declining profile, if not due to the anisotropy parameters or the incomplete membership, suggests a deviation from standard dark matter density profiles. We further identify three galaxy groups with unusually low dark matter fractions., Comment: 12 pages, 4 figures, published in MNRAS

This study proposes the use of Machine Learning models to predict the early onset of sepsis using deidentified clinical data from Montefiore Medical Center in Bronx, NY, USA. A supervised learning approach was adopted, wherein an XGBoost model was trained utilizing 80\% of the train dataset, encompassing 107 features (including the original and derived features). Subsequently, the model was evaluated on the remaining 20\% of the test data. The model was validated on prospective data that was entirely unseen during the training phase. To assess the model's performance at the individual patient level and timeliness of the prediction, a normalized utility score was employed, a widely recognized scoring methodology for sepsis detection, as outlined in the PhysioNet Sepsis Challenge paper. Metrics such as F1 Score, Sensitivity, Specificity, and Flag Rate were also devised. The model achieved a normalized utility score of 0.494 on test data and 0.378 on prospective data at threshold 0.3. The F1 scores were 80.8\% and 67.1\% respectively for the test data and the prospective data for the same threshold, highlighting its potential to be integrated into clinical decision-making processes effectively. These results bear testament to the model's robust predictive capabilities and its potential to substantially impact clinical decision-making processes., Comment: 16 pages, 6 figures and 7 tables