Search

As deep learning models gain attraction in medical data, ensuring transparent and trustworthy decision-making is essential. In skin cancer diagnosis, while advancements in lesion detection and classification have improved accuracy, the black-box nature of these methods poses challenges in understanding their decision processes, leading to trust issues among physicians. This study leverages the CLIP (Contrastive Language-Image Pretraining) model, trained on different skin lesion datasets, to capture meaningful relationships between visual features and diagnostic criteria terms. To further enhance transparency, we propose a method called MedGrad E-CLIP, which builds on gradient-based E-CLIP by incorporating a weighted entropy mechanism designed for complex medical imaging like skin lesions. This approach highlights critical image regions linked to specific diagnostic descriptions. The developed integrated pipeline not only classifies skin lesions by matching corresponding descriptions but also adds an essential layer of explainability developed especially for medical data. By visually explaining how different features in an image relates to diagnostic criteria, this approach demonstrates the potential of advanced vision-language models in medical image analysis, ultimately improving transparency, robustness, and trust in AI-driven diagnostic systems., Comment: Accepted to 2025 IEEE/CVF Winter Conference on Applications of Computer Vision Workshops (WACVW)

Bayesian inference is conceptually elegant, but calculating posterior expectations can entail a heavy computational cost. Monte Carlo methods are reliable and supported by strong asymptotic guarantees, but do not leverage smoothness of the integrand. Solving Stein equations has emerged as a possible alternative, providing a framework for numerical approximation of posterior expectations in which smoothness can be exploited. However, existing numerical methods for Stein equations are associated with high computational cost due to the need to solve large linear systems. This paper considers the combination of iterative linear solvers and preconditioning strategies to obtain fast approximate solutions of Stein equations. Our main finding is that these methods can be effective, but that the performance of different preconditioning strategies is context-dependent, so that no single strategy can be universally preferred.

We present results from a high cadence multi-wavelength observational campaign of the enigmatic changing look AGN 1ES 1927+654 from May 2022- April 2024, coincident with an unprecedented radio flare (an increase in flux by a factor of $\sim 60$ over a few months) and the emergence of a spatially resolved jet at $0.1-0.3$ pc scales (Meyer et al. 2024). Companion work has also detected a recurrent quasi-periodic oscillation (QPO) in the $2-10$ keV band with an increasing frequency ($1-2$ mHz) over the same period (Masterson et al., 2025). During this time, the soft X-rays ($0.3-2$ keV) monotonically increased by a factor of $\sim 8$, while the UV emission remained near-steady with $<30\%$ variation and the $2-10$ keV flux showed variation by a factor $\lesssim 2$. The weak variation of the $2-10$ keV X-ray emission and the stability of the UV emission suggest that the magnetic energy density and accretion rate are relatively unchanged, and that the jet could be launched due to a reconfiguration of the magnetic field (toroidal to poloidal) close to the black hole. Advecting poloidal flux onto the event horizon would trigger the Blandford-Znajek (BZ) mechanism, leading to the onset of the jet. The concurrent softening of the coronal slope (from $\Gamma= 2.70\pm 0.04$ to $\Gamma=3.27\pm 0.04$), the appearance of a QPO, and low coronal temperature ($kT_{e}=8_{-3}^{+8}$ keV) during the radio outburst suggest that the poloidal field reconfiguration can significantly impact coronal properties and thus influence jet dynamics. These extraordinary findings in real time are crucial for coronal and jet plasma studies, particularly as our results are independent of coronal geometry., Comment: Submitted to ApJ after minor referee comments

Robotic telescope networks play an important role in capturing early and bright optical afterglows, providing critical insights into the energetics and emission mechanisms of GRBs. In this study, we analyze GRB 230204B, an exceptionally energetic and multi-pulsed long GRB, detected by the Fermi GBM and MAXI detectors, with an isotropic equivalent gamma-ray energy exceeding 10$^{54}$ erg. Time-resolved spectral analysis reveals a transition in the prompt emission from hard (sub-photospheric dominated) spectra during early pulses to softer (synchrotron radiation dominated) spectra in later pulses, indicative of a hybrid jet composition. We report the discovery and characterization of the optical afterglow using the MASTER and BOOTES robotic telescope networks, alongside long-term radio observations extending to 335 days post-burst with the ATCA. At ~1.3 ks post-burst, the optical luminosity was exceptionally high, surpassing even other bright GRBs, such as GRB 221009A (the ``BOAT"). Multi-wavelength modeling, incorporating data from MASTER, BOOTES, DOT, Swift/XRT, and radio observations, was conducted using an external ISM forward-shock top-hat jet model with afterglowpy. The results reveal a narrow and highly collimated jet with a circumburst density of n$_{0}$ ~ 28.12 cm$^{-3}$, kinetic energy E$_{K}$ ~ 4.18 x 10$^{55}$ erg, and a relatively low value of $\epsilon_{B}$ = 2.14 x 10$^{-6}$, indicating shock-compression of the magnetic field in the surrounding interstellar medium. We constrained a low radiative efficiency of ~ 4.3 %. This study highlights the indispensable contribution of robotic networks to early afterglow observations and advances our understanding of GRB 230204B unique characteristics and underlying jet physics., Comment: 27 pages, 12 figures, 8 tables, submitted

The LIGO, Virgo and KAGRA gravitational wave observatories are currently undertaking their O4 observing run offering the opportunity to discover new electromagnetic counterparts to gravitational wave events. We examine the capability of the Neil Gehrels Swift Observatory (Swift) to respond to these triggers, primarily binary neutron star mergers, with both the UV/Optical Telescope (UVOT) and the X-ray Telescope (XRT). We simulate Swift's response to a trigger under different strategies using model skymaps, convolving these with the 2MPZ catalogue to produce an ordered list of observing fields, deriving the time taken for Swift to reach the correct field and simulating the instrumental responses to modelled kilonovae and short gamma-ray burst afterglows. We find that UVOT using the $u$ filter with an exposure time of order 120 s is optimal for most follow-up observations and that we are likely to detect counterparts in $\sim6$% of all binary neutron star triggers detectable by LVK in O4. We find that the gravitational wave 90% error area and measured distance to the trigger allow us to select optimal triggers to follow-up. Focussing on sources less than 300 Mpc away or 500 Mpc if the error area is less than a few hundred square degrees, distances greater than previously assumed, offer the best opportunity for discovery by Swift with $\sim5 - 30$% of triggers having detection probabilities $\geq 0.5$. At even greater distances, we can further optimise our follow-up by adopting a longer 250 s or 500 s exposure time., Comment: 16 pages, 13 figures. Final version accepted by MNRAS

The Legacy Survey of Space and Time (LSST) at Vera C. Rubin Observatory is planned to begin in the Fall of 2025. The LSST survey cadence has been designed via a community-driven process regulated by the Survey Cadence Optimization Committee (SCOC), which recommended up to 3% of the observing time to carry out Target of Opportunity (ToO) observations. Experts from the scientific community, Rubin Observatory personnel, and members of the SCOC were brought together to deliver a recommendation for the implementation of the ToO program during a workshop held in March 2024. Four main science cases were identified: gravitational wave multi-messenger astronomy, high energy neutrinos, Galactic supernovae, and small potentially hazardous asteroids possible impactors. Additional science cases were identified and briefly addressed in the documents, including lensed or poorly localized gamma-ray bursts and twilight discoveries. Trigger prioritization, automated response, and detailed strategies were discussed for each science case. This document represents the outcome of the Rubin ToO 2024 workshop, with additional contributions from members of the Rubin Science Collaborations. The implementation of the selection criteria and strategies presented in this document has been endorsed in the SCOC Phase 3 Recommendations document (PSTN-056). Although the ToO program is still to be finalized, this document serves as a baseline plan for ToO observations with the Rubin Observatory.

Vector Symbolic Architectures (VSAs) are one approach to developing Neuro-symbolic AI, where two vectors in $\mathbb{R}^d$ are `bound' together to produce a new vector in the same space. VSAs support the commutativity and associativity of this binding operation, along with an inverse operation, allowing one to construct symbolic-style manipulations over real-valued vectors. Most VSAs were developed before deep learning and automatic differentiation became popular and instead focused on efficacy in hand-designed systems. In this work, we introduce the Hadamard-derived linear Binding (HLB), which is designed to have favorable computational efficiency, and efficacy in classic VSA tasks, and perform well in differentiable systems. Code is available at https://github.com/FutureComputing4AI/Hadamard-derived-Linear-Binding, Comment: To appear in the 38th Conference on Neural Information Processing Systems (NeurIPS 2024)

Many students struggle with math word problems (MWPs), often finding it difficult to identify key information and select the appropriate mathematical operations. Schema-based instruction (SBI) is an evidence-based strategy that helps students categorize problems based on their structure, improving problem-solving accuracy. Building on this, we propose a Schema-Based Instruction Retrieval-Augmented Generation (SBI-RAG) framework that incorporates a large language model (LLM). Our approach emphasizes step-by-step reasoning by leveraging schemas to guide solution generation. We evaluate its performance on the GSM8K dataset, comparing it with GPT-4 and GPT-3.5 Turbo, and introduce a "reasoning score" metric to assess solution quality. Our findings suggest that SBI-RAG enhances reasoning clarity and facilitates a more structured problem-solving process potentially providing educational benefits for students., Comment: Accepted to the 4th MATH-AI Workshop at NeurIPS'24

Deterministic mathematical models, such as those specified via differential equations, are a powerful tool to communicate scientific insight. However, such models are necessarily simplified descriptions of the real world. Generalised Bayesian methodologies have been proposed for inference with misspecified models, but these are typically associated with vanishing parameter uncertainty as more data are observed. In the context of a misspecified deterministic mathematical model, this has the undesirable consequence that posterior predictions become deterministic and certain, while being incorrect. Taking this observation as a starting point, we propose Prediction-Centric Uncertainty Quantification, where a mixture distribution based on the deterministic model confers improved uncertainty quantification in the predictive context. Computation of the mixing distribution is cast as a (regularised) gradient flow of the maximum mean discrepancy (MMD), enabling consistent numerical approximations to be obtained. Results are reported on both a toy model from population ecology and a real model of protein signalling in cell biology.

This article appeared in the September 2024 issue (Vol. 31, No. 3) of the Bulletin of the International Society for Bayesian Analysis (ISBA).

SN 2023tsz is a Type Ibn supernova (SNe Ibn) discovered in an extremely low-mass host. SNe Ibn are an uncommon subtype of stripped-envelope core-collapse SNe. They are characterised by narrow helium emission lines in their spectra and are believed to originate from the collapse of massive Wolf-Rayet (WR) stars, though their progenitor systems still remain poorly understood. In terms of energetics and spectrophotometric evolution, SN 2023tsz is largely a typical example of the class, although line profile asymmetries in the nebular phase are seen, which may indicate the presence of dust formation or unshocked circumstellar material. Intriguingly, SN 2023tsz is located in an extraordinarily low-mass host galaxy that is in the 2nd percentile for SESN host masses and star formation rates (SFR). The host has a radius of 1.0 kpc, a $g$-band absolute magnitude of $-12.73$, and an estimated metallicity of $\log(Z_{*}/Z_{\odot}$) = $-1.56$. The SFR and metallicity of the host galaxy raise questions about the progenitor of SN 2023tsz. The low SFR suggests that a star with sufficient mass to evolve into a WR would be uncommon in this galaxy. Further, the very low-metallicity is a challenge for single stellar evolution to enable H and He stripping of the progenitor and produce a SN Ibn explosion. The host galaxy of SN 2023tsz adds another piece to the ongoing puzzle of SNe Ibn progenitors, and demonstrates that they can occur in hosts too faint to be observed in contemporary sky surveys at a more typical SN Ibn redshift., Comment: 14 pages, 10 figures, submitted to MNRAS

Quasi-periodic Eruptions (QPEs) are luminous bursts of soft X-rays from the nuclei of galaxies, repeating on timescales of hours to weeks. The mechanism behind these rare systems is uncertain, but most theories involve accretion disks around supermassive black holes (SMBHs), undergoing instabilities or interacting with a stellar object in a close orbit. It has been suggested that this disk could be created when the SMBH disrupts a passing star, implying that many QPEs should be preceded by observable tidal disruption events (TDEs). Two known QPE sources show long-term decays in quiescent luminosity consistent with TDEs, and two observed TDEs have exhibited X-ray flares consistent with individual eruptions. TDEs and QPEs also occur preferentially in similar galaxies. However, no confirmed repeating QPEs have been associated with a spectroscopically confirmed TDE or an optical TDE observed at peak brightness. Here we report the detection of nine X-ray QPEs with a mean recurrence time of approximately 48 hours from AT2019qiz, a nearby and extensively studied optically-selected TDE. We detect and model the X-ray, ultraviolet and optical emission from the accretion disk, and show that an orbiting body colliding with this disk provides a plausible explanation for the QPEs.

Notch signaling is a ubiquitous and versatile intercellular signaling system that drives collective behaviors and pattern formation in biological tissues. During embryonic development, Notch is involved in generation of collective biochemical oscillations that form the vertebrate body segments, and its failure results in embryonic defects. Notch ligands of the Delta family are key components of this collective rhythm, but it is unclear how different Delta ligands with distinct properties contribute to relaying information among cells. Motivated by the zebrafish segmentation clock, in this work we propose a theory describing interactions between biochemical oscillators, where Notch receptor is bound by both oscillatory and nonoscillatory Delta ligands. Based on previous in vitro binding studies, we first consider Notch activation by Delta dimers. This hypothesis is consistent with experimental observations in conditions of perturbed Notch signaling. Then we test an alternative hypothesis where Delta monomers directly bind and activate Notch, and show that this second model can also describe the experimental observations. We show that these two hypotheses assign different roles for a non-oscillatory ligand, as a binding partner or as a baseline signal. Finally, we discuss experiments to distinguish between the two scenarios. Broadly, this work highlights how a multiplicity of ligands may be harnessed by a signaling system to generate versatile responses.

This book aims to provide a graduate-level introduction to advanced topics in Markov chain Monte Carlo (MCMC) algorithms, as applied broadly in the Bayesian computational context. Most, if not all of these topics (stochastic gradient MCMC, non-reversible MCMC, continuous time MCMC, and new techniques for convergence assessment) have emerged as recently as the last decade, and have driven substantial recent practical and theoretical advances in the field. A particular focus is on methods that are scalable with respect to either the amount of data, or the data dimension, motivated by the emerging high-priority application areas in machine learning and AI., Comment: Preprint of upcoming book published by Cambridge University Press. Comments and feedback are welcome

We present multi-frequency (5-345 GHz) and multi-resolution radio observations of 1ES 1927+654, widely considered one of the most unusual and extreme changing-look active galactic nuclei (CL-AGN). The source was first designated a CL-AGN after an optical outburst in late 2017 and has since displayed considerable changes in X-ray emission, including the destruction and rebuilding of the X-ray corona in 2019-2020. Radio observations prior to 2023 show a faint and compact radio source typical of radio-quiet AGN. Starting in February 2023, 1ES 1927+654 began exhibiting a radio flare with a steep exponential rise, reaching a peak 60 times previous flux levels, and has maintained this higher level of radio emission for over a year to date. The 5-23 GHz spectrum is broadly similar to gigahertz-peaked radio sources, which are understood to be young radio jets less than ~1000 years old. Recent high-resolution VLBA observations at 23.5 GHz now show resolved extensions on either side of the core, with a separation of ~0.15 pc, consistent with a new and mildly relativistic bipolar outflow. A steady increase in the soft X-ray band (0.3-2 keV) concurrent with the radio may be consistent with jet-driven shocked gas, though further observations are needed to test alternate scenarios. This source joins a growing number of CL-AGN and tidal disruption events which show late-time radio activity, years after the initial outburst., Comment: Submitted to ApJ Letters 24 June 2024; Accepted 14 October 2024

We present a search for luminous, long-duration ambiguous nuclear transients (ANTs) similar to the unprecedented discovery of the extreme, ambiguous event AT2021lwx with a $>150$\,d rise time and luminosity $10^{45.7}$\,erg\,s$^{-1}$. We use the Lasair transient broker to search Zwicky Transient Facility (ZTF) data for transients lasting more than one year and exhibiting smooth declines. Our search returns 59 events, seven of which we classify as ANTs assumed to be driven by accretion onto supermassive black holes. We propose the remaining 52 are stochastic variability from regular supermassive black hole accretion rather than distinct transients. We supplement the seven ANTs with three nuclear transients in ZTF that fail the light curve selection but have clear single flares and spectra that do not resemble typical AGN. All of these 11 ANTs have a mid-infrared flare from an assumed dust echo, implying the ubiquity of dust around the black holes giving rise to ANTs. No events are more luminous than AT2021lwx, but one (ZTF19aamrjar) has twice the duration and a higher integrated energy release. On the other extreme, ZTF20abodaps reaches a luminosity close to AT2021lwx with a rise time $<20$\,d and that fades smoothly in $>600$\,d. We define a portion of rise-time versus flare amplitude space that selects ANTs with $\sim50$ per cent purity against variable active galactic nuclei. We calculate a volumetric rate of $\gtrsim 3\times10^{-11}$\,Mpc$^{-1}$\,yr$^{-1}$, consistent with the events being caused by tidal disruptions of intermediate and high-mass stars., Comment: 19 pages, 9 figures plus appendix. Accepted for publication in MNRAS. Data available at https://github.com/wisemanp/ANTs-Nest

Diffusion models are typically trained using score matching, yet score matching is agnostic to the particular forward process that defines the model. This paper argues that Markov diffusion models enjoy an advantage over other types of diffusion model, as their associated operators can be exploited to improve the training process. In particular, (i) there exists an explicit formal solution to the forward process as a sequence of time-dependent kernel mean embeddings; and (ii) the derivation of score-matching and related estimators can be streamlined. Building upon (i), we propose Riemannian diffusion kernel smoothing, which ameliorates the need for neural score approximation, at least in the low-dimensional context; Building upon (ii), we propose operator-informed score matching, a variance reduction technique that is straightforward to implement in both low- and high-dimensional diffusion modeling and is demonstrated to improve score matching in an empirical proof-of-concept., Comment: Preprint; 19 pages, 5 figures

The brightest Gamma-ray burst (GRB) ever, GRB 221009A, displays ultra-long GRB (ULGRB) characteristics, with a prompt emission duration exceeding 1000 s. To constrain the origin and central engine of this unique burst, we analyze its prompt and afterglow characteristics and compare them to the established set of similar GRBs. To achieve this, we statistically examine a nearly complete sample of Swift-detected GRBs with measured redshifts. Categorizing the sample to Bronze, Silver, and Gold by fitting a Gaussian function to the log-normal of T$_{90}$ duration distribution and considering three sub-samples respectively to 1, 2, and 3 times of the standard deviation to the mean value. GRB 221009A falls into the Gold sub-sample. Our analysis of prompt emission and afterglow characteristics aims to identify trends between the three burst groups. Notably, the Gold sub-sample (a higher likelihood of being ULGRB candidates) suggests a collapsar scenario with a hyper-accreting black hole as a potential central engine, while a few GRBs (GRB 060218, GRB 091024A, and GRB 100316D) in our Gold sub-sample favor a magnetar. Late-time near-IR (NIR) observations from 3.6m Devasthal Optical Telescope (DOT) rule out the presence of any bright supernova associated with GRB 221009A in the Gold sub-sample. To further constrain the physical properties of ULGRB progenitors, we employ the tool MESA to simulate the evolution of low-metallicity massive stars with different initial rotations. The outcomes suggest that rotating ($\Omega \geq 0.2\,\Omega_{\rm c}$) massive stars could potentially be the progenitors of ULGRBs within the considered parameters and initial inputs to MESA., Comment: Accepted in Astrophysical Journal: 39 pages, 19 Figures and 8 Tables

The multifaceted disruption of social and political processes during the most intense phases of the COVID crisis encouraged many groups and individuals to propose change and reform to education. This current report is a distinctive voice in the debate about 'assessment futures'. Not least, it outlines the nature of human construction of the future. It acknowledges that while the future might be constrained by current arrangements and the impact of past policy and practice, the future is not 'inevitable' or fixed, but is the product of human agency: a combination of thought, decisions, actions and reactions. The aim of this research, looking forward to 2050, was not to predict the future of assessment but to use Futures Thinking as an approach to anticipate plausible future directions. The research was carried out in collaboration with individuals and groups from various departments across the University of Cambridge and two external groups. The participants took part in three 3-hour long workshops, with many completing significant work in-between the first and the second workshop. This study highlighted fundamental distinctions between linear and nonlinear approaches to future forecasting. It emphasises the necessity for unconventional methodologies to anticipate discontinuities and surprises resulting from radical shifts and disruptive innovation.

Most stripped-envelope supernova progenitors are thought to be formed through binary interaction, losing hydrogen and/or helium from their outer layers. Ultra-stripped supernovae are an emerging class of transient which are expected to be produced through envelope-stripping by a NS companion. However, relatively few examples are known and the outcomes of such systems can be diverse and are poorly understood at present. Here, we present spectroscopic observations and high-cadence, multi-band photometry of SN 2023zaw, a rapidly evolving supernova with a low ejecta mass discovered in a nearby spiral galaxy at D = 39.7 Mpc. It has significant Milky Way extinction, $E(B-V)_{\rm MW} = 0.21$, and significant (but uncertain) host extinction. Bayesian evidence comparison reveals that nickel is not the only power source and an additional energy source is required to explain our observations. Our models suggest an ejecta mass of $M_{\rm ej} \sim 0.07$ $\rm M_{\odot}$ and a synthesized nickel mass of $M_{\rm Ni} \sim 0.007$ $\rm M_{\odot}$ are required to explain the observations. We find that additional heating from a central engine, or interaction with circumstellar material can power the early light curve., Comment: Accepted for publication in ApJL

An informal observation, made by several authors, is that the adaptive design of a Markov transition kernel has the flavour of a reinforcement learning task. Yet, to-date it has remained unclear how to actually exploit modern reinforcement learning technologies for adaptive MCMC. The aim of this paper is to set out a general framework, called Reinforcement Learning Metropolis--Hastings, that is theoretically supported and empirically validated. Our principal focus is on learning fast-mixing Metropolis--Hastings transition kernels, which we cast as deterministic policies and optimise via a policy gradient. Control of the learning rate provably ensures conditions for ergodicity are satisfied. The methodology is used to construct a gradient-free sampler that out-performs a popular gradient-free adaptive Metropolis--Hastings algorithm on $\approx 90 \%$ of tasks in the PosteriorDB benchmark.

We present the largest optical photometry compilation of Gamma-Ray Bursts (GRBs) with redshifts ($z$). We include 64813 observations of 535 events (including upper limits) from 28 February 1997 up to 18 August 2023. We also present a user-friendly web tool \textit{grbLC} which allows users the visualization of photometry, coordinates, redshift, host galaxy extinction, and spectral indices for each event in our database. Furthermore, we have added a Gamma Ray Coordinate Network (GCN) scraper that can be used to collect data by gathering magnitudes from the GCNs. The web tool also includes a package for uniformly investigating colour evolution. We compute the optical spectral indices for 138 GRBs for which we have at least 4 filters at the same epoch in our sample and craft a procedure to distinguish between GRBs with and without colour evolution. By providing a uniform format and repository for the optical catalogue, this web-based archive is the first step towards unifying several community efforts to gather the photometric information for all GRBs with known redshifts. This catalogue will enable population studies by providing light curves (LCs) with better coverage since we have gathered data from different ground-based locations. Consequently, these LCs can be used to train future LC reconstructions for an extended inference of the redshift. The data gathering also allows us to fill some of the orbital gaps from Swift in crucial points of the LCs, e.g., at the end of the plateau emission or where a jet break is identified., Comment: 20 pages, 16 figures, 2 tables. Submitted to MNRAS, this version matches the third revision. The Online Materials and data will be available after the publication

The nature of the minute-to-hour long Fast X-ray Transients (FXTs) localised by telescopes such as Chandra, Swift, and XMM-Newton remains mysterious, with numerous models suggested for the events. Here, we report multi-wavelength observations of EP240315a, a 1600 s long transient detected by the Einstein Probe, showing it to have a redshift of z=4.859. We measure a low column density of neutral hydrogen, indicating that the event is embedded in a low-density environment, further supported by direct detection of leaking ionising Lyman-continuum. The observed properties are consistent with EP240315a being a long-duration gamma-ray burst, and these observations support an interpretation in which a significant fraction of the FXT population are lower-luminosity examples of similar events. Such transients are detectable at high redshifts by the Einstein Probe and, in the (near) future, out to even larger distances by SVOM, THESEUS, and Athena, providing samples of events into the epoch of reionisation., Comment: 41 pages, 7 figures, submitted

Malware detection is an interesting and valuable domain to work in because it has significant real-world impact and unique machine-learning challenges. We investigate existing long-range techniques and benchmarks and find that they're not very suitable in this problem area. In this paper, we introduce Holographic Global Convolutional Networks (HGConv) that utilize the properties of Holographic Reduced Representations (HRR) to encode and decode features from sequence elements. Unlike other global convolutional methods, our method does not require any intricate kernel computation or crafted kernel design. HGConv kernels are defined as simple parameters learned through backpropagation. The proposed method has achieved new SOTA results on Microsoft Malware Classification Challenge, Drebin, and EMBER malware benchmarks. With log-linear complexity in sequence length, the empirical results demonstrate substantially faster run-time by HGConv compared to other methods achieving far more efficient scaling even with sequence length $\geq 100,000$., Comment: To appear in Proceedings of the 27th International Conference on Artificial Intelligence and Statistics (AISTATS) 2024, Valencia, Spain

In the rapidly evolving landscape of communication and network security, the increasing reliance on deep neural networks (DNNs) and cloud services for data processing presents a significant vulnerability: the potential for backdoors that can be exploited by malicious actors. Our approach leverages advanced tensor decomposition algorithms Independent Vector Analysis (IVA), Multiset Canonical Correlation Analysis (MCCA), and Parallel Factor Analysis (PARAFAC2) to meticulously analyze the weights of pre-trained DNNs and distinguish between backdoored and clean models effectively. The key strengths of our method lie in its domain independence, adaptability to various network architectures, and ability to operate without access to the training data of the scrutinized models. This not only ensures versatility across different application scenarios but also addresses the challenge of identifying backdoors without prior knowledge of the specific triggers employed to alter network behavior. We have applied our detection pipeline to three distinct computer vision datasets, encompassing both image classification and object detection tasks. The results demonstrate a marked improvement in both accuracy and efficiency over existing backdoor detection methods. This advancement enhances the security of deep learning and AI in networked systems, providing essential cybersecurity against evolving threats in emerging technologies., Comment: 6 pages, Accepted at the International Conference on Communications 2024. arXiv admin note: text overlap with arXiv:2212.08121

Gamma-ray bursts (GRBs) are ideal probes of the Universe at high redshift (z > 5), pinpointing the locations of the earliest star-forming galaxies and providing bright backlights that can be used to spectrally fingerprint the intergalactic medium and host galaxy during the period of reionization. Future missions such as Gamow Explorer are being proposed to unlock this potential by increasing the rate of identification of high-z GRBs to rapidly trigger observations from 6-10 m ground telescopes, JWST, and the Extremely Large Telescopes. Gamow was proposed to the NASA 2021 Medium-Class Explorer (MIDEX) program as a fast-slewing satellite featuring a wide-field lobster-eye X-ray telescope (LEXT) to detect and localize GRBs, and a 30 cm narrow-field multi-channel photo-z infrared telescope (PIRT) to measure their photometric redshifts using the Lyman-alpha dropout technique. To derive the PIRT sensitivity requirement we compiled a complete sample of GRB optical-near-infrared afterglows from 2008 to 2021, adding a total of 66 new afterglows to our earlier sample, including all known high-z GRB afterglows. We performed full light-curve and spectral-energy-distribution analyses of these afterglows to derive their true luminosity at very early times. For all the light curves, where possible, we determined the brightness at the time of the initial finding chart of Gamow, at different high redshifts and in different NIR bands. We then followed the evolution of the luminosity to predict requirements for ground and space-based follow-up. We find that a PIRT sensitivity of 15 micro-Jy (21 mag AB) in a 500 s exposure simultaneously in five NIR bands within 1000s of the GRB trigger will meet the Gamow mission requirement to recover > 80% of all redshifts at z > 5., Comment: 44 pages, 10 figures, 1 table. Accepted for publication in Astronomy and Astrophysics 15 Feb 2024. Abstract abridged for arXiv

This paper gives a detailed account of the analysis underpinning the 2021 update to the list of standard reference frequency values recommended by the International Committee for Weights and Measures (CIPM). This update focused on a subset of atomic transitions that are secondary representations of the second (SRS) or considered as potential SRS. As in previous updates in 2015 and 2017, methods for analysing over-determined data sets were applied to make optimum use of the worldwide body of published clock comparison data. To ensure that these methods were robust, three independent calculations were performed using two different algorithms. The 2021 update differed from previous updates in taking detailed account of correlations among the input data, a step shown to be important in deriving unbiased frequency values and avoiding underestimation of their uncertainties. It also differed in the procedures used to assess input data and to assign uncertainties to the recommended frequency values, with previous practice being adapted to produce a fully consistent output data set consisting of frequency ratio values as well as absolute frequencies. These changes are significant in the context of an anticipated redefinition of the second in terms of an optical transition or transitions, since optical frequency ratio measurements will be critical for verifying the international consistency of optical clocks prior to the redefinition. In the meantime, the reduced uncertainties for optical SRS resulting from this analysis significantly increases the weight that secondary frequency standards based on these transitions can have in the steering of International Atomic Time (TAI)., Comment: 22 pages, 5 figures

As deep neural networks and the datasets used to train them get larger, the default approach to integrating them into research and commercial projects is to download a pre-trained model and fine tune it. But these models can have uncertain provenance, opening up the possibility that they embed hidden malicious behavior such as trojans or backdoors, where small changes to an input (triggers) can cause the model to produce incorrect outputs (e.g., to misclassify). This paper introduces a novel approach to backdoor detection that uses two tensor decomposition methods applied to network activations. This has a number of advantages relative to existing detection methods, including the ability to analyze multiple models at the same time, working across a wide variety of network architectures, making no assumptions about the nature of triggers used to alter network behavior, and being computationally efficient. We provide a detailed description of the detection pipeline along with results on models trained on the MNIST digit dataset, CIFAR-10 dataset, and two difficult datasets from NIST's TrojAI competition. These results show that our method detects backdoored networks more accurately and efficiently than current state-of-the-art methods.

While deep learning has enjoyed significant success in computer vision tasks over the past decade, many shortcomings still exist from a Cognitive Science (CogSci) perspective. In particular, the ability to subitize, i.e., quickly and accurately identify the small (less than 6) count of items, is not well learned by current Convolutional Neural Networks (CNNs) or Vision Transformers (ViTs) when using a standard cross-entropy (CE) loss. In this paper, we demonstrate that adapting tools used in CogSci research can improve the subitizing generalization of CNNs and ViTs by developing an alternative loss function using Holographic Reduced Representations (HRRs). We investigate how this neuro-symbolic approach to learning affects the subitizing capability of CNNs and ViTs, and so we focus on specially crafted problems that isolate generalization to specific aspects of subitizing. Via saliency maps and out-of-distribution performance, we are able to empirically observe that the proposed HRR loss improves subitizing generalization though it does not completely solve the problem. In addition, we find that ViTs perform considerably worse compared to CNNs in most respects on subitizing, except on one axis where an HRR-based loss provides improvement., Comment: Accepted in 38th Annual AAAI Workshop on Neuro-Symbolic Learning and Reasoning in the Era of Large Language Models (NuCLeaR), 2024

Background. Large Language Models (LLMs) hold promise for improving genetic variant literature review in clinical testing. We assessed Generative Pretrained Transformer 4's (GPT-4) performance, nondeterminism, and drift to inform its suitability for use in complex clinical processes. Methods. A 2-prompt process for classification of functional evidence was optimized using a development set of 45 articles. The prompts asked GPT-4 to supply all functional data present in an article related to a variant or indicate that no functional evidence is present. For articles indicated as containing functional evidence, a second prompt asked GPT-4 to classify the evidence into pathogenic, benign, or intermediate/inconclusive categories. A final test set of 72 manually classified articles was used to test performance. Results. Over a 2.5-month period (Dec 2023-Feb 2024), we observed substantial differences in intraday (nondeterminism) and across day (drift) results, which lessened after 1/18/24. This variability is seen within and across models in the GPT-4 series, affecting different performance statistics to different degrees. Twenty runs after 1/18/24 identified articles containing functional evidence with 92.2% sensitivity, 95.6% positive predictive value (PPV) and 86.3% negative predictive value (NPV). The second prompt's identified pathogenic functional evidence with 90.0% sensitivity, 74.0% PPV and 95.3% NVP and for benign evidence with 88.0% sensitivity, 76.6% PPV and 96.9% NVP. Conclusion. Nondeterminism and drift within LLMs must be assessed and monitored when introducing LLM based functionality into clinical workflows. Failing to do this assessment or accounting for these challenges could lead to incorrect or missing information that is critical for patient care. The performance of our prompts appears adequate to assist in article prioritization but not in automated decision making., Comment: 5 pages, 1 table, 4 figures, 2 supplementary tables, 1 supplementary figure. These authors contributed equally: Samuel J. Aronson, Kalotina Machini, and Jiyeon Shin Corresponding author: Samuel J. Aronson

Differential Diagnosis (DDx) is the process of identifying the most likely medical condition among the possible pathologies through the process of elimination based on evidence. An automated process that narrows a large set of pathologies down to the most likely pathologies will be of great importance. The primary prior works have relied on the Reinforcement Learning (RL) paradigm under the intuition that it aligns better with how physicians perform DDx. In this paper, we show that a generative approach trained with simpler supervised and self-supervised learning signals can achieve superior results on the current benchmark. The proposed Transformer-based generative network, named DDxT, autoregressively produces a set of possible pathologies, i.e., DDx, and predicts the actual pathology using a neural network. Experiments are performed using the DDXPlus dataset. In the case of DDx, the proposed network has achieved a mean accuracy of 99.82% and a mean F1 score of 0.9472. Additionally, mean accuracy reaches 99.98% with a mean F1 score of 0.9949 while predicting ground truth pathology. The proposed DDxT outperformed the previous RL-based approaches by a big margin. Overall, the automated Transformer-based DDx generative model has the potential to become a useful tool for a physician in times of urgency., Comment: Accepted at 1st Workshop on Deep Generative Models for Health at NeurIPS 2023