Search

Current ton-scale direct detection experiments have started observing solar neutrinos. In this paper, we probe the weak mixing angle using the latest direct detection data. Utilizing the recent measurement of $^8$B solar neutrinos through coherent neutrino-nucleus scattering by PandaX-4T, we demonstrate that it can probe the weak mixing angle in a complementary region with an error bar comparable to that of dedicated neutrino experiments. Additionally, we show that the current XENONnT electron recoil data can probe the weak mixing angle through neutrino-electron scattering. This occurs in a momentum transfer region that is more than an order of magnitude smaller than the region probed by atomic parity violation experiments. Our findings show huge scope of probing a Standard Model parameter in an entirely new energy regime through the observation of neutrinos in future direct detection experiments., Comment: 7 pages, 2 figures. Comments are welcome

While scaling training compute has led to remarkable improvements in large language models (LLMs), scaling inference compute has not yet yielded analogous gains. We hypothesize that a core missing component is a lack of diverse LLM outputs, leading to inefficient search due to models repeatedly sampling highly similar, yet incorrect generations. We empirically demonstrate that this lack of diversity can be mitigated by searching over candidate plans for solving a problem in natural language. Based on this insight, we propose PLANSEARCH, a novel search algorithm which shows strong results across HumanEval+, MBPP+, and LiveCodeBench (a contamination-free benchmark for competitive coding). PLANSEARCH generates a diverse set of observations about the problem and then uses these observations to construct plans for solving the problem. By searching over plans in natural language rather than directly over code solutions, PLANSEARCH explores a significantly more diverse range of potential solutions compared to baseline search methods. Using PLANSEARCH on top of Claude 3.5 Sonnet achieves a state-of-the-art pass@200 of 77.0% on LiveCodeBench, outperforming both the best score achieved without search (pass@1 = 41.4%) and using standard repeated sampling (pass@200 = 60.6%). Finally, we show that, across all models, search algorithms, and benchmarks analyzed, we can accurately predict performance gains due to search as a direct function of the diversity over generated ideas.

Quantum complexity has emerged as a central concept in diverse areas of physics, ranging from quantum computing to the theory of black holes. We perform a systematic study of complexity in random quantum circuits with and without measurements. We observe that complexity grows linearly before saturating to a constant value. For $N$ qubits without measurements, the saturation value scales as $2^{N-1}$, and the saturation time scales as $2^N$. This behaviour remains identical in the presence of random measurements with different probabilities, indicating that this notion of complexity is insensitive to the rate of measurement. We also study the behaviour of complexity in two variants of the random unitary floquet circuit, where we observe that complexity acts as a novel probe of Anderson localization and many-body localization., Comment: 6 pages, 3 figures

The interplay of magnetic correlations, crystal electric field interactions, and spin-orbit coupling in low-dimensional frustrated magnets fosters novel ground states with unusual excitations. Here, we report the magnetic properties and crystal electric field (CEF) scheme of a rare-earth-based square-lattice antiferromagnet NdKNaNbO$_5$ investigated via magnetization, specific heat, electron spin resonance (ESR), and inelastic neutron scattering (INS) experiments. The low-temperature Curie-Weiss temperature $\theta_{\rm CW} \simeq -0.6$ K implies net antiferromagnetic interactions between the Nd$^{3+}$ ions. Two broad maxima are observed in the low temperature specific heat data in magnetic fields, indicating multilevel Schottky anomalies due to the effect of CEF. No magnetic long-range-order is detected down to 0.4 K. The CEF excitations of Kramers' ion Nd$^{3+}$ ($J=9/2$) probed via INS experiments evince dispersionless excitations characterizing the transitions among the CEF energy levels. The fit of the INS spectra enabled the mapping of the CEF Hamiltonian and the energy eigenvalues of the Kramers' doublets. The simulation using the obtained CEF parameters reproduces the broad maxima in specific heat in zero-field as well as in different applied fields. The significant contribution from $J_z = \pm 1/2$ state to the wave function of the ground state doublet indicates the role of strong quantum fluctuations at low temperatures. The magnetic ground state is found to be a Kramers' doublet with effective spin $J_{\rm eff} = 1/2$ at low temperatures., Comment: 14 pages, 10 figures

This study explores integrating reinforcement learning (RL) with idealised climate models to address key parameterisation challenges in climate science. Current climate models rely on complex mathematical parameterisations to represent sub-grid scale processes, which can introduce substantial uncertainties. RL offers capabilities to enhance these parameterisation schemes, including direct interaction, handling sparse or delayed feedback, continuous online learning, and long-term optimisation. We evaluate the performance of eight RL algorithms on two idealised environments: one for temperature bias correction, another for radiative-convective equilibrium (RCE) imitating real-world computational constraints. Results show different RL approaches excel in different climate scenarios with exploration algorithms performing better in bias correction, while exploitation algorithms proving more effective for RCE. These findings support the potential of RL-based parameterisation schemes to be integrated into global climate models, improving accuracy and efficiency in capturing complex climate dynamics. Overall, this work represents an important first step towards leveraging RL to enhance climate model accuracy, critical for improving climate understanding and predictions. Code accessible at https://github.com/p3jitnath/climate-rl., Comment: 24 pages, 9 figures

Knowledge Distillation (KD) is a powerful approach for compressing a large model into a smaller, more efficient model, particularly beneficial for latency-sensitive applications like recommender systems. However, current KD research predominantly focuses on Computer Vision (CV) and NLP tasks, overlooking unique data characteristics and challenges inherent to recommender systems. This paper addresses these overlooked challenges, specifically: (1) mitigating data distribution shifts between teacher and student models, (2) efficiently identifying optimal teacher configurations within time and budgetary constraints, and (3) enabling computationally efficient and rapid sharing of teacher labels to support multiple students. We present a robust KD system developed and rigorously evaluated on multiple large-scale personalized video recommendation systems within Google. Our live experiment results demonstrate significant improvements in student model performance while ensuring consistent and reliable generation of high quality teacher labels from a continuous data stream of data.

The recent detection of large column density absorption lines from highly ionized gas in a few directions through the circumgalactic medium (CGM) of the Milky Way (MW) has been puzzling. The inferred temperature from these absorption lines far exceeds the virial temperature of the MW, and the column densities are also too large to be easily explained. In this paper, we propose a novel idea to explain these observations and claim that they may not have originated from the CGM, but from a totally different type of source, namely, stellar ejecta from supernovae (SNe) above the Galactic disk that happen to lie in the line of sight to the background quasars. About $\sim 20\%$ of massive OB stars (progenitors of core-collapse supernovae) are known to be runaway stars that have high ejection velocities near the Galactic plane and can end up exploding as SNe above the Galactic disk. We show that the associated reverse shock in the supernova remnant in the early non-radiative phase can heat the ejecta to temperatures of $\gtrsim 10^7\,{\rm K}$ and can naturally explain the observed high column density of ions in the observed `super-virial' phase along with $\alpha$-enriched super-solar abundance that is typical of core-collapse supernovae. However, SNe from runaway stars has a covering fraction of $\lesssim 0.7 \%$ and thus can only explain the observations along limited sightlines., Comment: 14 pages, submitted to the Astrophysical Journal (ApJ)

The presence of the $\approx 10^6$ K gas in the circumgalactic medium of the Milky Way has been well established. However, the location and the origin of the newly discovered hot gas at `super-virial' temperatures of $\approx 10^7$ K have been puzzling. This hot gas has been detected in both absorption and emission; here we focus on the emitting gas only. We show that both the `virial' and the `super-virial' temperature gas as observed in \emph{emission} occupy disk-like extraplanar regions, in addition to the diffuse virial temperature gas filling the halo of the Milky Way. We perform idealized hydrodynamical simulations to show that the $\approx 10^7$ K emitting gas is likely to be produced by stellar feedback in and around the Galactic disk. We further show that the emitting gas at both super-virial and virial temperatures in the extraplanar regions is metal enriched and is not in hydrostatic equilibrium with the halo but is continuously evolving., Comment: 11 pages, accepted for publication in the Astrophysical Journal (ApJ)

Let $pod_2(n)$ denote the number of $2$-regular partitions of $n$ with distinct odd parts (even parts are unrestricted). In this article, we obtain congruences for $pod_2(n)$ mod $2$ and mod $8$ using some generating function manipulations and the theory of Hecke eigenform., Comment: 8 pages

Since the release of Segment Anything 2 (SAM2), the medical imaging community has been actively evaluating its performance for 3D medical image segmentation. However, different studies have employed varying evaluation pipelines, resulting in conflicting outcomes that obscure a clear understanding of SAM2's capabilities and potential applications. We shortly review existing benchmarks and point out that the SAM2 paper clearly outlines a zero-shot evaluation pipeline, which simulates user clicks iteratively for up to eight iterations. We reproduced this interactive annotation simulation on 3D CT datasets and provided the results and code~\url{https://github.com/Project-MONAI/VISTA}. Our findings reveal that directly applying SAM2 on 3D medical imaging in a zero-shot manner is far from satisfactory. It is prone to generating false positives when foreground objects disappear, and annotating more slices cannot fully offset this tendency. For smaller single-connected objects like kidney and aorta, SAM2 performs reasonably well but for most organs it is still far behind state-of-the-art 3D annotation methods. More research and innovation are needed for 3D medical imaging community to use SAM2 correctly.

Optical networks play a crucial role in todays digital topography, enabling the high-speed and reliable transmission of vast amounts of data over optical fibre for long distances. This paper provides an overview of optical networks, especially emphasising on their evolution with time.

To address the data scarcity associated with 3D assets, 2D-lifting techniques such as Score Distillation Sampling (SDS) have become a widely adopted practice in text-to-3D generation pipelines. However, the diffusion models used in these techniques are prone to viewpoint bias and thus lead to geometric inconsistencies such as the Janus problem. To counter this, we introduce MT3D, a text-to-3D generative model that leverages a high-fidelity 3D object to overcome viewpoint bias and explicitly infuse geometric understanding into the generation pipeline. Firstly, we employ depth maps derived from a high-quality 3D model as control signals to guarantee that the generated 2D images preserve the fundamental shape and structure, thereby reducing the inherent viewpoint bias. Next, we utilize deep geometric moments to ensure geometric consistency in the 3D representation explicitly. By incorporating geometric details from a 3D asset, MT3D enables the creation of diverse and geometrically consistent objects, thereby improving the quality and usability of our 3D representations., Comment: 9 pages, 8 figures

Sequences of the form $(\gcd(u_n,v_n))_{n \in \mathbb N}$, with $(u_n)_n$, $(v_n)_n$ sums of $S$-units, have been considered by several authors. The study of $\gcd(n,u_n)$ corresponds, following Silverman, to divisibility sequences arising from the split algebraic group $\mathbb G_{\mathrm{a}} \times \mathbb G_{\mathrm{m}}$; in this case, Sanna determined all asymptotic moments of the arithmetic function $\log\,\gcd (n,u_n)$ when $(u_n)_n$ is a Lucas sequence. Here, we characterize the asymptotic behavior of the moments themselves $\sum_{n \leq x}\,\gcd(n,u_n)^\lambda$, thus solving the moment problem for $\mathbb G_{\mathrm{a}} \times \mathbb G_{\mathrm{m}}$. We give both unconditional and conditional results, the latter only relying on standard conjectures in analytic number theory., Comment: 15 pages, 1 figure

Let $G$ be a group. Associate a graph $\mathcal{E}_G$ (called the co-Engel graph of $G$) with $G$ whose vertex set is $G$ and two distinct vertices $x$ and $y$ are adjacent if $[x, {}_k y] \neq 1$ and $[y, {}_k x] \neq 1$ for all positive integer $k$. This graph, under the name ``Engel graph'', was introduced by Abdollahi. Let $L(G)$ be the set of all left Engel elements of $G$. In this paper, we realize the induced subgraph of co-Engel graphs of certain finite non-Engel groups $G$ induced by $G \setminus L(G)$. We write $\mathcal{E}^-(G)$ to denote the subgraph of $\mathcal{E}_G$ induced by $G \setminus L(G)$. We also compute genus, various spectra, energies and Zagreb indices of $\mathcal{E}^-(G)$ for those groups. As a consequence, we determine (up to isomorphism) all finite non-Engel group $G$ such that the clique number is at most $4$ and $\mathcal{E}^-$ is toroidal or projective. Further, we show that $\coeng{G}$ is super integral and satisfies the E-LE conjecture and the Hansen--Vuki{\v{c}}evi{\'c} conjecture for the groups considered in this paper.

Let A be a graph type and B an equivalence relation on a group $G$. Let $[g]$ be the equivalence class of $g$ with respect to the equivalence relation B. The B superA graph of $G$ is an undirected graph whose vertex set is $G$ and two distinct vertices $g, h \in G$ are adjacent if $[g] = [h]$ or there exist $x \in [g]$ and $y \in [h]$ such that $x$ and $y$ are adjacent in the A graph of $G$. In this paper, we compute spectrum of equality/conjugacy supercommuting graphs of dihedral/dicyclic groups and show that these graphs are not integral.

Purification maps a mixed state to a pure state and a non-unitary evolution into a unitary one by enlarging the Hilbert space. We link the operator complexity of the density matrix to the state/operator complexity of purified states using three purification schemes: time-independent, time-dependent, and instantaneous purification. We propose inequalities among the operator and state complexities of mixed states and their purifications, demonstrated with a single qubit, two-qubit Werner states, and infinite-dimensional diagonal mixed states. We find that the complexity of a vacuum evolving into a thermal state equals the average number of Rindler particles created between left and right Rindler wedges. Finally, for the thermofield double state evolving from zero to finite temperature, we show that 1) the state complexity follows the Lloyd bound, reminiscent of the quantum speed limit, and 2) the Krylov state/operator complexities are subadditive in contrast to the holographic volume complexity., Comment: 11 pages, 5 figures; Lloyd bound with the energy is added and a slight rearrangement of texts, typos are corrected, references are added

A matrix $A\in \mathbb{R}^{m \times n}$ is strictly sign regular/SSR (or sign regular/SR) if for each $1 \leq k \leq \min \{ m, n \}$, all $k\times k$ minors of $A$ (or non-zero $k\times k$ minors of $A$) have the same sign. This class of matrices contains the totally positive matrices, and was first studied by Schoenberg (1930) to characterize Variation Diminution (VD), a fundamental property in total positivity theory. In this note, we classify all surjective linear mappings $\mathcal{L}:\mathbb{R}^{m\times n}\to\mathbb{R}^{m\times n}$ that preserve: (i) sign regularity and (ii) sign regularity with a given sign pattern, as well as (iii) strict versions of these., Comment: This is the second part of the paper arXiv:2307.11822v2, which is now split in into two parts. The first part is accepted; this is the second part, which is moreover extensively revised (22 pages)

Motivated by the recent Belle II result indicating a $2.7\,\sigma$ excess of $B^+\to K^+ + E_{\text{miss}}$ events compared to the Standard Model (SM) prediction for $B^+ \to K^+ \nu\bar{\nu}$, we explore an explanation to this anomaly based on a KSVZ-type QCD axion model featuring a Peccei-Quinn (PQ) symmetry breaking at high-scale, that can provide a solution to the strong CP problem with dark matter relic abundance. The model contains a PQ-charged scalar} leptoquark which can interact with the SM quarks only via mass-mixing of the latter with vector-like quarks. The mixing between SM and vector-like quarks is determined by the PQ mass scales and can explain the excess $B^+\to K^+ + E_{\text{miss}}$ events while respecting other flavour constraints. The same PQ-charged scalar leptoquarks and vector-like quarks also mediate the two-loop radiative neutrino masses., Comment: 25 pages, 6 figures

We analyze the properties of Krylov state complexity in qubit dynamics, considering a single qubit and a qubit pair. A geometrical picture of the Krylov complexity is discussed for the single-qubit case, whereas it becomes non-trivial for the two-qubit case. Considering the particular case of interacting Rydberg two-level atoms, we show that the Krylov basis obtained using an effective Hamiltonian minimizes the complexity compared to that which is obtained from the original Hamiltonian. We further generalize the latter property to an arbitrary Hamiltonian in which the entire Hilbert space comprises of two subspaces provided a weak coupling between them., Comment: 14 pages, 8 figures

We investigate the ground state phases of an extended Bose-Hubbard ladder of unit filling via the density-matrix-renormalization-group method and, in particular, the effect of rung-hoppings. In contrast to a single-chain, a commensurate supersolid emerges, and based on the Luttinger parameter, we classify them into two types. The latter leads to a reentrant gapless behavior as the onsite interaction is increased while keeping all other parameters intact. A reentrant gapped transition is also found as a function of nearest-neighbor interactions. Further, we show that the string order characterizing the Haldane phase vanishes for a finite inter-chain hopping amplitude, however small it is. Finally, we propose two experimental platforms to observe our findings, using either dipolar atoms or polar molecules and Rydberg admixed atoms., Comment: 7 pages, 9 figures

In the context of the CERN EP R&D on monolithic sensors and the ALICE ITS3 upgrade, the Tower Partners Semiconductor Co (TPSCo) 65 nm process has been qualified for use in high energy physics, and adopted for the ALICE ITS3 upgrade. An Analog Pixel Test Structure (APTS) featuring fast per pixel operational-amplifier-based buffering for a small matrix of four by four pixels, with a sensor with a small collection electrode and a very non-uniform electric field, was designed to allow detailed characterization of the pixel performance in this technology. Several variants of this chip with different pixel designs have been characterized with a (120 GeV/$c$) positive hadron beam. This result indicates that the APTS-OA prototype variants with the best performance achieve a time resolution of 63 ps with a detection efficiency exceeding 99% and a spatial resolution of 2 $\mu$m, highlighting the potential of TPSCo 65nm CMOS imaging technology for high-energy physics and other fields requiring precise time measurement, high detection efficiency, and excellent spatial resolution.

Techniques that learn improved representations via offline data or self-supervised objectives have shown impressive results in traditional reinforcement learning (RL). Nevertheless, it is unclear how improved representation learning can benefit reinforcement learning from human feedback (RLHF) on language models (LMs). In this work, we propose training reward models (RMs) in a contrastive, $\textit{goal-conditioned}$ fashion by increasing the representation similarity of future states along sampled preferred trajectories and decreasing the similarity along randomly sampled dispreferred trajectories. This objective significantly improves RM performance by up to 0.09 AUROC across challenging benchmarks, such as MATH and GSM8k. These findings extend to general alignment as well -- on the Helpful-Harmless dataset, we observe $2.3\%$ increase in accuracy. Beyond improving reward model performance, we show this way of training RM representations enables improved $\textit{steerability}$ because it allows us to evaluate the likelihood of an action achieving a particular goal-state (e.g., whether a solution is correct or helpful). Leveraging this insight, we find that we can filter up to $55\%$ of generated tokens during majority voting by discarding trajectories likely to end up in an "incorrect" state, which leads to significant cost savings. We additionally find that these representations can perform fine-grained control by conditioning on desired future goal-states. For example, we show that steering a Llama 3 model towards helpful generations with our approach improves helpfulness by $9.6\%$ over a supervised-fine-tuning trained baseline. Similarly, steering the model towards complex generations improves complexity by $21.6\%$ over the baseline. Overall, we find that training RMs in this contrastive, goal-conditioned fashion significantly improves performance and enables model steerability.

Vector maps find widespread utility across diverse domains due to their capacity to not only store but also represent discrete data boundaries such as building footprints, disaster impact analysis, digitization, urban planning, location points, transport links, and more. Although extensive research exists on identifying building footprints and road types from satellite imagery, the generation of vector maps from such imagery remains an area with limited exploration. Furthermore, conventional map generation techniques rely on labor-intensive manual feature extraction or rule-based approaches, which impose inherent limitations. To surmount these limitations, we propose a novel method called HPix, which utilizes modified Generative Adversarial Networks (GANs) to generate vector tile map from satellite images. HPix incorporates two hierarchical frameworks: one operating at the global level and the other at the local level, resulting in a comprehensive model. Through empirical evaluations, our proposed approach showcases its effectiveness in producing highly accurate and visually captivating vector tile maps derived from satellite images. We further extend our study's application to include mapping of road intersections and building footprints cluster based on their area.

Continuum equations are ubiquitous in physical modelling of elastic, viscous, and viscoelastic systems. The equations of continuum mechanics take nontrivial forms on curved surfaces. Although the curved surface formulation of the continuum equations are derived in many excellent references available in the literature, they are not readily usable for solving physical problems due to the covariant, contravariant or mixed nature of the stress and strain tensors in the equations. We present the continuum equations in terms of physical components in a general differentiable manifold. This general formulation of the continuum equations can be used readily for modelling physical problems on arbitrary curved surfaces. We demonstrate this with the help of some examples., Comment: Only the correct affiliation string and the funding acknowledgement has been added in the newer version

Quasi-normal modes of NSs are an exciting prospect for analyzing the internal composition of NSs and studying matter at high densities. In this work, we focus on studying the $f$- and $p$- quadrupolar oscillation modes, which couple with gravitational waves. We construct two different EOS ensembles, one without and one with a first-order phase transition, and examine how $f$- and $p$-modes might help us differentiate them. We find ensemble specific exclusion regions in the $65\%$ and $95\%$ confidence contours of the frequency-damping time relations. The exclusion regions become more prominent for the higher-order oscillation modes. However, these modes have higher frequencies, which are beyond the detection capabilities of present gravitational wave detectors. The quasi-universal relations of dimensionless quantities prove to be ineffective in differentiating the EOS ensembles, as they obscure the details of the EOS., Comment: 12 pages, 7 figures

This paper introduces nonlinear fractional Lane-Emden equations of the form, $$ D^{\alpha} y(x) + \frac{\lambda}{x^\beta}~ D^{\beta} y(x) + f(y) =0, ~ ~1 < \alpha \leq 2, ~~ 0< \beta \leq 1, ~~ 0 < x < 1,$$ subject to boundary conditions, $$ y'(0) =\mathbf{a} , ~~~ \mathbf{c}~ y'(1) + \mathbf{d}~ y(1) = \mathbf{b},$$ where, $D^\alpha, D^\beta$ represent Caputo fractional derivative, $\mathbf{a, b,c,d} \in \mathbb{R}$, $ \lambda = 1, 2$, and $f(y)$ is non linear function of $y.$ We have developed collocation method namely, uniform fractional Haar wavelet collocation method and used it to compute solutions. The proposed method combines the quasilinearization method with the Haar wavelet collocation method. In this approach, fractional Haar integrations is used to determine the linear system, which, upon solving, produces the required solution. Our findings suggest that as the values of $(\alpha, \beta)$ approach $(2,1),$ the solutions of the fractional and classical Lane-Emden become identical.

The rapid growth in the use of Large Language Models (LLMs) and AI Agents as part of software development and deployment is revolutionizing the information technology landscape. While code generation receives significant attention, a higher-impact application lies in using AI agents for operational resilience of cloud services, which currently require significant human effort and domain knowledge. There is a growing interest in AI for IT Operations (AIOps) which aims to automate complex operational tasks, like fault localization and root cause analysis, thereby reducing human intervention and customer impact. However, achieving the vision of autonomous and self-healing clouds through AIOps is hampered by the lack of standardized frameworks for building, evaluating, and improving AIOps agents. This vision paper lays the groundwork for such a framework by first framing the requirements and then discussing design decisions that satisfy them. We also propose AIOpsLab, a prototype implementation leveraging agent-cloud-interface that orchestrates an application, injects real-time faults using chaos engineering, and interfaces with an agent to localize and resolve the faults. We report promising results and lay the groundwork to build a modular and robust framework for building, evaluating, and improving agents for autonomous clouds.

In this work, we propose a management method for controlling the speed and direction of self-bound quantum droplets (QDs) in a binary Bose-Einstein condensate mixture under time-modulated external harmonic confinement. Utilizing the 1D extended Gross-Pit"{a}evskii equation, QDs are constructed within both regular and expulsive parabolic traps, considering temporally varying attractive quadratic beyond mean field and repulsive cubic mean-field atom-atom interactions. Through the derived wavefunction solution, we illustrate the dynamics of slowing, stopping, reversing, fragmentation, collapse, and revival of droplets. Additionally, the solutions reveal a crystalline order with a superfluid background, indicative of supersolid behavior in various parameter domains. Notably, one-third of the constant background matches the lowest residual condensate. These findings hold potential applications in matter-wave interferometry and quantum information processing.

Let $B$ be an equivalence relation defined on a finite group $G$. The $B$ super commuting graph on $G$ is a graph whose vertex set is $G$ and two distinct vertices $g$ and $h$ are adjacent if either $[g] = [h]$ or there exist $g' \in [g]$ and $h' \in [h]$ such that $g'$ commutes with $h'$, where $[g]$ is the $B$-equivalence class of $g \in G$. Considering $B$ as the equality, conjugacy and same order relations on $G$, in this article, we discuss the graph structures of equality/conjugacy/order super commuting graphs of certain well-known families of non-abelian groups viz. dihedral groups, dicyclic groups, semidihedral groups, quasidihedral groups, the groups $U_{6n}, V_{8n}, M_{2mn}$ etc. Further, we compute the Zagreb indices of these graphs and show that they satisfy Hansen-Vuki{\v{c}}evi{\'c} conjecture., Comment: 21 pages

In this paper, we consider the set of partitions $pend(n)$ which enumerates the number of partitions of $n$ wherein the even parts are not allowed to be distinct. Using a result of Newman, we prove a few infinite families of congruences modulo 2 for $pend(n)$., Comment: 6 pages

We characterize quantum dynamics in triangular billiards in terms of five properties: (1) the level spacing ratio (LSR), (2) spectral complexity (SC), (3) Lanczos coefficient variance, (4) energy eigenstate localisation in the Krylov basis, and (5) dynamical growth of spread complexity. The billiards we study are classified as integrable, pseudointegrable or non-integrable, depending on their internal angles which determine properties of classical trajectories and associated quantum spectral statistics. A consistent picture emerges when transitioning from integrable to non-integrable triangles: (1) LSRs increase; (2) spectral complexity growth slows down; (3) Lanczos coefficient variances decrease; (4) energy eigenstates delocalize in the Krylov basis; and (5) spread complexity increases, displaying a peak prior to a plateau instead of recurrences. Pseudo-integrable triangles deviate by a small amount in these charactertistics from non-integrable ones, which in turn approximate models from the Gaussian Orthogonal Ensemble (GOE). Isosceles pseudointegrable and non-integrable triangles have independent sectors that are symmetric and antisymmetric under a reflection symmetry. These sectors separately reproduce characteristics of the GOE, even though the combined system approximates characteristics expected from integrable theories with Poisson distributed spectra., Comment: 25 pages, 12 figures

The determination of the direction of a stellar core collapse via its neutrino emission is crucial for the identification of the progenitor for a multimessenger follow-up. A highly effective method of reconstructing supernova directions within the Deep Underground Neutrino Experiment (DUNE) is introduced. The supernova neutrino pointing resolution is studied by simulating and reconstructing electron-neutrino charged-current absorption on $^{40}$Ar and elastic scattering of neutrinos on electrons. Procedures to reconstruct individual interactions, including a newly developed technique called ``brems flipping'', as well as the burst direction from an ensemble of interactions are described. Performance of the burst direction reconstruction is evaluated for supernovae happening at a distance of 10 kpc for a specific supernova burst flux model. The pointing resolution is found to be 3.4 degrees at 68% coverage for a perfect interaction-channel classification and a fiducial mass of 40 kton, and 6.6 degrees for a 10 kton fiducial mass respectively. Assuming a 4% rate of charged-current interactions being misidentified as elastic scattering, DUNE's burst pointing resolution is found to be 4.3 degrees (8.7 degrees) at 68% coverage., Comment: 25 pages, 16 figures

The layered triangular lattice owing to $1:2$ order of $B$ and $B'$ sites in the triple perovskite $A_3 B B'_2$O$_9$ family provides an enticing domain for exploring the complex phenomena of quantum spin liquids (QSLs). We report a comprehensive investigation of the ground state properties of Sr$_3$CuTa$_2$O$_9$ that belongs to the above family, by employing magnetization, specific heat, and muon spin relaxation ($\mu$SR) experiments down to the lowest temperature of 0.1~K. Analysis of the magnetic susceptibility indicates that the spin-lattice is a nearly isotropic $S = 1/2$ triangular lattice. We illustrate the observation of a gapless QSL, in which conventional spin ordering or freezing effects are absent, even at temperatures more than two orders of magnitude smaller than the exchange energy ($J_{\rm CW}/k_{\rm B} \simeq -5.04$~K). Magnetic specific heat in zero-field follows a power law, $C_{\rm m} \sim T^\eta$, below 1.2~K with $\eta \approx 2/3$, which is consistent with a theoretical proposal of the presence of spinon Fermi surface. Below 1.2~K, the $\mu$SR relaxation rate shows no temperature dependence, suggesting persistent spin dynamics as expected for a QSL state. Delving deeper, we also analyze longitudinal field $\mu$SR spectra revealing strong dynamical correlations in the spin-disordered ground state. All of these highlight the characteristics of spin entanglement in the QSL state., Comment: 7 pages, 4 figures. Accepted for publication at Physical Review B (Letter)

Trillium lattices formed by corner-shared triangular units are the platform for magnetic frustration in three dimensions. Herein, we report structural and magnetic properties of the Cr-based double trillium lattice material KBaCr$_2$(PO$_4$)$_3$ studied by x-ray diffraction, magnetization, heat capacity, thermal conductivity, and $^{31}$P nuclear magnetic resonance (NMR) measurements complemented by density-functional band-structure calculations. Heat capacity and $^{31}$P NMR measurements reveal the magnetic transition at $T_{\rm N1} \simeq 13.5$ K in zero field followed by another transition at $T_{\rm N2} \simeq 7$ K in weak applied fields. The NMR sublattice magnetization confirms that the transition at $T_{\rm N1}$ is 3D in nature. The $^{31}$P spin-lattice relaxation rate in the ordered state follows the $T^3$ behavior indicative of the two-magnon Raman process. The spin lattice of KBaCr$_2$(PO$_4$)$_3$ comprises two crystallographically nonequivalent ferromagnetic sublattices that are coupled antiferromagnetically, thus eliminating frustration in this trillium network., Comment: 12 pages, 13 figures

Digital imaging systems have classically been based on brute-force measuring and processing of pixels organized on regular grids. The human visual system, on the other hand, performs a massive data reduction from the number of photo-receptors to the optic nerve, essentially encoding the image information into a low bandwidth latent space representation suitable for processing by the human brain. In this work, we propose to follow a similar approach for the development of artificial vision systems. Latent Space Imaging is a new paradigm that, through a combination of optics and software, directly encodes the image information into the semantically rich latent space of a generative model, thus substantially reducing bandwidth and memory requirements during the capture process. We demonstrate this new principle through an initial hardware prototype based on the single pixel camera. By designing an amplitude modulation scheme that encodes into the latent space of a generative model, we achieve compression ratios from 1:100 to 1:1,000 during the imaging process, illustrating the potential of latent space imaging for highly efficient imaging hardware, to enable future applications in high speed imaging, or task-specific cameras with substantially reduced hardware complexity.

We have proposed an arbitrary vector beam (VB) generation scheme in a thin disk-shaped quantum dot (QD) medium considering phonon interaction. The QD biexciton system exhibits interplay between first and third-order nonlinear susceptibility between two orthogonal circular polarisation transitions. Three QD transitions are coupled with one applied weak and two strong control orbital angular momentum (OAM) carrying fields. Therefore, the applied field experiences absorption, and a new field with the desired OAM is generated via four-wave mixing (FWM). These two orthogonal field superpositions produce VB at the QD medium end. We have also demonstrated the polarization rotation of a VB by changing only the relative control field phase. Additionally, we have analyzed the effect of temperature on the VB generation., Comment: 10 pages, 5 figures

This paper, for the first time, marries large foundation models with human sketch understanding. We demonstrate what this brings -- a paradigm shift in terms of generalised sketch representation learning (e.g., classification). This generalisation happens on two fronts: (i) generalisation across unknown categories (i.e., open-set), and (ii) generalisation traversing abstraction levels (i.e., good and bad sketches), both being timely challenges that remain unsolved in the sketch literature. Our design is intuitive and centred around transferring the already stellar generalisation ability of CLIP to benefit generalised learning for sketches. We first "condition" the vanilla CLIP model by learning sketch-specific prompts using a novel auxiliary head of raster to vector sketch conversion. This importantly makes CLIP "sketch-aware". We then make CLIP acute to the inherently different sketch abstraction levels. This is achieved by learning a codebook of abstraction-specific prompt biases, a weighted combination of which facilitates the representation of sketches across abstraction levels -- low abstract edge-maps, medium abstract sketches in TU-Berlin, and highly abstract doodles in QuickDraw. Our framework surpasses popular sketch representation learning algorithms in both zero-shot and few-shot setups and in novel settings across different abstraction boundaries., Comment: ECCV 2024

In digital pathology, the traditional method for deep learning-based image segmentation typically involves a two-stage process: initially segmenting high-resolution whole slide images (WSI) into smaller patches (e.g., 256x256, 512x512, 1024x1024) and subsequently reconstructing them to their original scale. This method often struggles to capture the complex details and vast scope of WSIs. In this paper, we propose the holistic histopathology (HoloHisto) segmentation method to achieve end-to-end segmentation on gigapixel WSIs, whose maximum resolution is above 80,000$\times$70,000 pixels. HoloHisto fundamentally shifts the paradigm of WSI segmentation to an end-to-end learning fashion with 1) a large (4K) resolution base patch for elevated visual information inclusion and efficient processing, and 2) a novel sequential tokenization mechanism to properly model the contextual relationships and efficiently model the rich information from the 4K input. To our best knowledge, HoloHisto presents the first holistic approach for gigapixel resolution WSI segmentation, supporting direct I/O of complete WSI and their corresponding gigapixel masks. Under the HoloHisto platform, we unveil a random 4K sampler that transcends ultra-high resolution, delivering 31 and 10 times more pixels than standard 2D and 3D patches, respectively, for advancing computational capabilities. To facilitate efficient 4K resolution dense prediction, we leverage sequential tokenization, utilizing a pre-trained image tokenizer to group image features into a discrete token grid. To assess the performance, our team curated a new kidney pathology image segmentation (KPIs) dataset with WSI-level glomeruli segmentation from whole mouse kidneys. From the results, HoloHisto-4K delivers remarkable performance gains over previous state-of-the-art models.

Large vision language models (VLMs) have progressed incredibly from research to applicability for general-purpose use cases. LLaVA-Med, a pioneering large language and vision assistant for biomedicine, can perform multi-modal biomedical image and data analysis to provide a natural language interface for radiologists. While it is highly generalizable and works with multi-modal data, it is currently limited by well-known challenges that exist in the large language model space. Hallucinations and imprecision in responses can lead to misdiagnosis which currently hinder the clinical adaptability of VLMs. To create precise, user-friendly models in healthcare, we propose D-Rax -- a domain-specific, conversational, radiologic assistance tool that can be used to gain insights about a particular radiologic image. In this study, we enhance the conversational analysis of chest X-ray (CXR) images to support radiological reporting, offering comprehensive insights from medical imaging and aiding in the formulation of accurate diagnosis. D-Rax is achieved by fine-tuning the LLaVA-Med architecture on our curated enhanced instruction-following data, comprising of images, instructions, as well as disease diagnosis and demographic predictions derived from MIMIC-CXR imaging data, CXR-related visual question answer (VQA) pairs, and predictive outcomes from multiple expert AI models. We observe statistically significant improvement in responses when evaluated for both open and close-ended conversations. Leveraging the power of state-of-the-art diagnostic models combined with VLMs, D-Rax empowers clinicians to interact with medical images using natural language, which could potentially streamline their decision-making process, enhance diagnostic accuracy, and conserve their time., Comment: accepted to the MICCAI 2024 Second International Workshop on Foundation Models for General Medical AI

One of the fundamental interactions that is found in many functional materials is the spin-phonon coupling (SPC), which is at the heart of many novel functionalities. The simultaneous presence of multi-magnetic phases makes SPC even more intriguing. We have used Raman spectroscopy as well as first-principles methods to investigate the possibility of the appearance of SPC in Gd-doped Y$_2$CoMnO$_6$ (YGCMO) double perovskite oxide and the influence of anti-site disorder on the same. YGCMO is found to exhibit anti-site disorder leading to both ferromagnetic (between Co and Mn) and anti-ferromagnetic interactions (Co-Co, Mn-Mn, Gd-Co/Mn). An analysis of the temperature-dependent phonon frequency for the stretching modes of YGCMO, obtained using RAMAN spectroscopy, indicates that SPC is possibly emerging from simultaneous presence of ferromagnetic and antiferromagnetic interactions. The nature of the phonon linewidth and the insulating state of the material eliminate the role of magnetostriction on the observed anomaly. The spin-phonon coupling strength comes out to be 0.29 cm$^{-1}$. Our experimental findings are corroborated by first-principles DFT calculations which indicate the presence of SPC in ordered YGCMO getting enhanced in the presence of anti-site disorder. This indicates a strong influence of B-site (Co/Mn) ordering on SPC in the bulk double perovskite systems. An analysis of the cause behind the enhanced SPC in the presence of anti-site disorder is also presented., Comment: 23 pages, 14 figures

In this paper, we delve into the intricate dynamics of Fine-Grained Sketch-Based Image Retrieval (FG-SBIR) by addressing a critical yet overlooked aspect -- the choice of viewpoint during sketch creation. Unlike photo systems that seamlessly handle diverse views through extensive datasets, sketch systems, with limited data collected from fixed perspectives, face challenges. Our pilot study, employing a pre-trained FG-SBIR model, highlights the system's struggle when query-sketches differ in viewpoint from target instances. Interestingly, a questionnaire however shows users desire autonomy, with a significant percentage favouring view-specific retrieval. To reconcile this, we advocate for a view-aware system, seamlessly accommodating both view-agnostic and view-specific tasks. Overcoming dataset limitations, our first contribution leverages multi-view 2D projections of 3D objects, instilling cross-modal view awareness. The second contribution introduces a customisable cross-modal feature through disentanglement, allowing effortless mode switching. Extensive experiments on standard datasets validate the effectiveness of our method., Comment: Accepted in European Conference on Computer Vision (ECCV) 2024

Opinion summarization in e-commerce encapsulates the collective views of numerous users about a product based on their reviews. Typically, a product on an e-commerce platform has thousands of reviews, each review comprising around 10-15 words. While Large Language Models (LLMs) have shown proficiency in summarization tasks, they struggle to handle such a large volume of reviews due to context limitations. To mitigate, we propose a scalable framework called Xl-OpSumm that generates summaries incrementally. However, the existing test set, AMASUM has only 560 reviews per product on average. Due to the lack of a test set with thousands of reviews, we created a new test set called Xl-Flipkart by gathering data from the Flipkart website and generating summaries using GPT-4. Through various automatic evaluations and extensive analysis, we evaluated the framework's efficiency on two datasets, AMASUM and Xl-Flipkart. Experimental results show that our framework, Xl-OpSumm powered by Llama-3-8B-8k, achieves an average ROUGE-1 F1 gain of 4.38% and a ROUGE-L F1 gain of 3.70% over the next best-performing model.

The description of dark matter as a pressure-less fluid and of dark energy as a cosmological constant, both minimally coupled to gravity, constitutes the basis of the concordance $\Lambda\text{CDM}$ model. However, the concordance model is based on using equations of motion directly for the fluids with constraints placed on their sources, and lacks an underlying Lagrangian. In this work, we propose a Lagrangian model of two spin zero fields describing dark energy and dark matter with an interaction term between the two along with self-interactions. We study the background evolution of the fields as well as their linear perturbations, suggesting an alternative to $\Lambda$CDM with dark matter and dark energy being fundamental dynamical fields. The parameters of the model are extracted using a Bayesian inference tool based on multiple cosmological data sets which include those of Planck (with lensing), BAO, Pantheon, SH0ES, and WiggleZ. Using these data, we set constraints on the dark matter mass and the interaction strengths. Furthermore, we find that the model is able to alleviate the Hubble tension for some data sets while also resolving the $S_8$ tension., Comment: 41 pages, 10 figures and 2 tables

A detailed study of the structural and magnetic properties of the spin-$3/2$ hyperkagome lattice compound Li$_2$MgMn$_3$O$_8$ is reported. This material shows ferromagnetic response below $T_{\rm C} \simeq 20.6$ K, the temperature almost three times lower than the Curie-Weiss temperature $\theta_{\rm CW} \simeq 56.6$ K. Density-functional band-structure calculations suggest that this reduction in $T_{\rm C}$ may be caused by long-range antiferromagnetic couplings that frustrate nearest-neighbor ferromagnetic couplings on the hyperkagome lattice. Large magnetocaloric effect is observed around the $T_{\rm C}$ with a maximum value of isothermal entropy change $\Delta S_{\rm m}\simeq 20$ J/kg-K and a maximum relative cooling power of $RCP\simeq 840$ J/kg for the 7 T magnetic field change. Critical analysis of the magnetization data and scaling analysis of the magnetocaloric effect suggest the 3D Heisenberg/XY universality class of the transition. The DC and AC magnetization measurements further reveal glassy nature of the ferromagnetic transition. A detailed study of the non-equilibrium dynamics via magnetic relaxation and memory effect measurements demonstrates that the system evolves through a large number of intermediate metastable states and manifests significant memory effect in the cluster-glass state., Comment: 17 pages, 14 figures

We report results from a detailed study of the neutron star X-ray binary, 4U 1608-52 using observations with {\it AstroSat} (LAXPC/SXT) and {\it NICER} during its 2016 and 2020 outbursts. The 0.7--20.0 keV spectra could be well described with the disk blackbody and thermal Comptonization model. The best-fitting inner disk temperature is $\sim$ 1 keV and radius { $\sim$ 22.17$^{+2.57}_{-2.38}$--27.19$^{+2.03}_{-1.85}$} km and no significant evolution was observed in the disk radius after performing flux and time-resolved spectroscopy. We used a multi-Lorentzian approach to fit the power density spectra and obtained broad-band noise variability. We estimated the energy-dependent fractional root mean square and time-lag of the broad-band noise, and these variations are quantitatively modelled as being due to the coherent variation of the disk emission and the coronal heating rate. Thus, the rapid temporal modeling is consistent with the longer term spectral evolution where the inner disk radius does not vary, and instead the variations can be attributed to accretion rate variations which changes the inner disk temperature and the coronal heating rate., Comment: 17 pages, 13 figures, The manuscript has been accepted for publication in The Astrophysical Journal

Recently, there has been much work on the design of general heuristics for graph-based, combinatorial optimization problems via the incorporation of Graph Neural Networks (GNNs) to learn distribution-specific solution structures.However, there is a lack of consistency in the evaluation of these heuristics, in terms of the baselines and instances chosen, which makes it difficult to assess the relative performance of the algorithms. In this paper, we propose an open-source benchmark suite MaxCut-Bench dedicated to the NP-hard Maximum Cut problem in both its weighted and unweighted variants, based on a careful selection of instances curated from diverse graph datasets. The suite offers a unified interface to various heuristics, both traditional and machine learning-based. Next, we use the benchmark in an attempt to systematically corroborate or reproduce the results of several, popular learning-based approaches, including S2V-DQN [31], ECO-DQN [4], among others, in terms of three dimensions: objective value, generalization, and scalability. Our empirical results show that several of the learned heuristics fail to outperform a naive greedy algorithm, and that only one of them consistently outperforms Tabu Search, a simple, general heuristic based upon local search. Furthermore, we find that the performance of ECO-DQN remains the same or is improved if the GNN is replaced by a simple linear regression on a subset of the features that are related to Tabu Search. Code, data, and pretrained models are available at: \url{https://github.com/ankurnath/MaxCut-Bench}.

In this short note, we revisit the entanglement wedge representation of AdS$_3$ bulk fields in terms of CFT operator product expansion (OPE) blocks for a general class of blocks. Given a boundary interval and its associated causal diamond, the OPEs involve boundary operators with or without spin, and located either at spacelike or timelike edges of the diamond. Only for a subset of these cases, can the OPE block be dual to a geodesic bulk field. We show that when applied to de Sitter, a suitable combination of Euclidean OPE blocks can represent a dS scalar integrated over the timelike extremal surfaces, which play an important role in defining pseudo-entropy. We also work out some simple higher dimensional examples., Comment: Version 1. 37 pages, 2 figures

A two-coil wearable system is proposed for wireless communication and powering between a transmitter coil in a necklace and a receiver coil in a smart contact lens, where the necklace is invisible in contrast to coils embedded in wearables like spectacles or headbands. Magneto-quasistatic(MQS) field coupling facilitates communication between the transmitter in the necklace and the contact lens receiver, enabling AR/VR and health monitoring. As long as the receiver coil remains within the magnetic field generated by the transmitter, continuous communication is sustained through MQS field coupling despite the misalignments present. Resonant frequency tuning enhances system efficiency. The system's performance was tested for coil misalignments, showing a maximum path loss variation within $10 dB$ across scenarios, indicating robustness. Finite Element Method(FEM) analysis has been used to study the system for efficient wireless data transfer and powering. A communication channel capacity is $4.5 Mbps$ over a $1 MHz$ bandwidth. Simulations show negligible path loss differences with or without human tissues, as magnetic coupling remains unaffected at MQS frequencies below $30 MHz$ due to similar magnetic permeability of tissues and air. Therefore, the possibility of efficient communication and powering of smart contact lenses through a necklace is shown for the first time using resonant MQS coupling at an axial distance of $15cm$ and lateral distance of over $9cm$ to enable AR/VR and health monitoring on the contact lens., Comment: 9 pages, 12 figures