Search

The fall 2023 and spring 2024 undergraduate enrollment increases, marking the first growth since the COVID-19 pandemic, show signs of a post-pandemic turnaround for higher education. However, a significant share of current undergraduates will eventually disengage from college before earning a degree or other credential. They will join tens of millions of other adult Americans who are Some College, No Credential (SCNC). The SCNC population has been consistently rising over time. Re-engaging those who stop out remains a persistent challenge and a priority for the forty states that have set ambitious postsecondary attainment goals. This report aims to provide timely insights into the SCNC population, offering state leaders and policymakers accurate data on its current status, along with tracking progress and outcome measures for SCNC students. The first section of this report describes who makes up the SCNC population and how it has changed since the last report. In this section, the authors pay particular attention to Recent Stopouts, who joined the SCNC population after being stopped out between January 2021 and July 2022. In the second section, the authors report on SCNC re-enrollment in the 2022-23 academic year as well as first-year credential earning for re-enrollees. The authors also provide new updates on continued enrollment and second-year credential earning for SCNC re-enrollees in the 2021-22 academic year, whom were first reported on last year.

Human-like personality traits have recently been discovered in large language models, raising the hypothesis that their (known and as yet undiscovered) biases conform with human latent psychological constructs. While large conversational models may be tricked into answering psychometric questionnaires, the latent psychological constructs of thousands of simpler transformers, trained for other tasks, cannot be assessed because appropriate psychometric methods are currently lacking. Here, we show how standard psychological questionnaires can be reformulated into natural language inference prompts, and we provide a code library to support the psychometric assessment of arbitrary models. We demonstrate, using a sample of 88 publicly available models, the existence of human-like mental health-related constructs (including anxiety, depression, and Sense of Coherence) which conform with standard theories in human psychology and show similar correlations and mitigation strategies. The ability to interpret and rectify the performance of language models by using psychological tools can boost the development of more explainable, controllable, and trustworthy models.

We propose a variant of chain of thought (CoT) prompting called Program Trace Prompting that makes explanations more observable while preserving the power, generality and flexibility of CoT. In our approach, few-shot CoT demonstrations are wrapped in a formal syntax based on Python, and each prompt: identifies and names steps; defines the input/output behavior of steps; and replaces CoT explanations of in-context examples with chains of these formalized steps on the same examples. Program Trace Prompting is applicable to many tasks, achieving strong results on the 23 diverse tasks in the BIG-Bench Hard benchmark. More importantly, by instrumenting explanations in this way, we enable new types of analysis. In particular, we identify "non-local errors" (which correspond to incorrectly learning the reasoning method illustrated in the demonstrations) as an unaddressed issue in CoT learning, and we present methods for verifying the modularity of steps in a CoT explanation.

We present Jamba-1.5, new instruction-tuned large language models based on our Jamba architecture. Jamba is a hybrid Transformer-Mamba mixture of experts architecture, providing high throughput and low memory usage across context lengths, while retaining the same or better quality as Transformer models. We release two model sizes: Jamba-1.5-Large, with 94B active parameters, and Jamba-1.5-Mini, with 12B active parameters. Both models are fine-tuned for a variety of conversational and instruction-following capabilties, and have an effective context length of 256K tokens, the largest amongst open-weight models. To support cost-effective inference, we introduce ExpertsInt8, a novel quantization technique that allows fitting Jamba-1.5-Large on a machine with 8 80GB GPUs when processing 256K-token contexts without loss of quality. When evaluated on a battery of academic and chatbot benchmarks, Jamba-1.5 models achieve excellent results while providing high throughput and outperforming other open-weight models on long-context benchmarks. The model weights for both sizes are publicly available under the Jamba Open Model License and we release ExpertsInt8 as open source., Comment: Webpage: https://www.ai21.com/jamba

Assessing instruction quality is a fundamental component of any improvement efforts in the education system. However, traditional manual assessments are expensive, subjective, and heavily dependent on observers' expertise and idiosyncratic factors, preventing teachers from getting timely and frequent feedback. Different from prior research that focuses on low-inference instructional practices, this paper presents the first study that leverages Natural Language Processing (NLP) techniques to assess multiple high-inference instructional practices in two distinct educational settings: in-person K-12 classrooms and simulated performance tasks for pre-service teachers. This is also the first study that applies NLP to measure a teaching practice that has been demonstrated to be particularly effective for students with special needs. We confront two challenges inherent in NLP-based instructional analysis, including noisy and long input data and highly skewed distributions of human ratings. Our results suggest that pretrained Language Models (PLMs) demonstrate performances comparable to the agreement level of human raters for variables that are more discrete and require lower inference, but their efficacy diminishes with more complex teaching practices. Interestingly, using only teachers' utterances as input yields strong results for student-centered variables, alleviating common concerns over the difficulty of collecting and transcribing high-quality student speech data in in-person teaching settings. Our findings highlight both the potential and the limitations of current NLP techniques in the education domain, opening avenues for further exploration.

This First Look report provides selected findings from the High School Longitudinal Study of 2009 (HSLS:09) Postsecondary Education Administrative Records Collection (PEAR). HSLS:09 follows a nationally representative sample of students who were ninth-graders in fall 2009 from high school into postsecondary education and the workforce. The PEAR data collection was conducted in 2021, approximately 8 years after high school graduation for most of the cohort. These data provide information on whether fall 2009 ninth-graders enrolled in postsecondary education by June 2021, and allow researchers to examine enrollment characteristics, degree completion, and financial aid awards for the subset of fall 2009 ninth-graders who enrolled in postsecondary education.

Despite calls to modernize education preparation, the way we train, support, and grow educators has remained largely unchanged for decades. But some educator training and professional development organizations are taking a different approach by offering educators flexible and job-embedded learning opportunities that recognize and validate learning through demonstrations of competence. Educators earn "micro-credentials" in the form of digital badges, which capture both the skill the educator demonstrated and the evidence they used to prove their mastery of that skill. This case study offers a look at one micro-credential program, developed by UCLA's ExcEL Leadership Academy that has been approved for ESOL teacher certification in Rhode Island. The program offers a progression of 12 micro-credentials focused on the skills and competencies educators need to serve multilingual learners (MLLs) effectively. Additionally, the case study offers recommendations for other states that hope to offer their educators high-quality competency-based pathways to certification and/or professional growth.

Information retrieval (IR) methods, like retrieval augmented generation, are fundamental to modern applications but often lack statistical guarantees. Conformal prediction addresses this by retrieving sets guaranteed to include relevant information, yet existing approaches produce large-sized sets, incurring high computational costs and slow response times. In this work, we introduce a score refinement method that applies a simple monotone transformation to retrieval scores, leading to significantly smaller conformal sets while maintaining their statistical guarantees. Experiments on various BEIR benchmarks validate the effectiveness of our approach in producing compact sets containing relevant information., Comment: 6 pages

Diacritical marks in the Hebrew language give words their vocalized form. The task of adding diacritical marks to plain Hebrew text is still dominated by a system that relies heavily on human-curated resources. Recent models trained on diacritized Hebrew texts still present a gap in performance. We use a recently developed char-based PLM to narrowly bridge this gap. Presenting MenakBERT, a character level transformer pretrained on Hebrew text and fine-tuned to produce diacritical marks for Hebrew sentences. We continue to show how finetuning a model for diacritizing transfers to a task such as part of speech tagging., Comment: Published at ISCOL2022 as a poster

We revisit the recently introduced Local Glivenko-Cantelli setting, which studies distribution-dependent uniform convegence rates of the Maximum Likelihood Estimator (MLE). In this work, we investigate generalizations of this setting where arbitrary estimators are allowed rather than just the MLE. Can a strictly larger class of measures be learned? Can better risk decay rates be obtained? We provide exhaustive answers to these questions -- which are both negative, provided the learner is barred from exploiting some infinite-dimensional pathologies. On the other hand, allowing such exploits does lead to a strictly larger class of learnable measures.

The practical use of text-to-image generation has evolved from simple, monolithic models to complex workflows that combine multiple specialized components. While workflow-based approaches can lead to improved image quality, crafting effective workflows requires significant expertise, owing to the large number of available components, their complex inter-dependence, and their dependence on the generation prompt. Here, we introduce the novel task of prompt-adaptive workflow generation, where the goal is to automatically tailor a workflow to each user prompt. We propose two LLM-based approaches to tackle this task: a tuning-based method that learns from user-preference data, and a training-free method that uses the LLM to select existing flows. Both approaches lead to improved image quality when compared to monolithic models or generic, prompt-independent workflows. Our work shows that prompt-dependent flow prediction offers a new pathway to improving text-to-image generation quality, complementing existing research directions in the field., Comment: Project website: https://comfygen-paper.github.io/

Large language models (LLMs) deployed as agents solve user-specified tasks over multiple steps while keeping the required manual engagement to a minimum. Crucially, such LLMs need to ground their generations in any feedback obtained to reliably achieve desired outcomes. We propose an end-to-end reinforcement learning method for teaching models to leverage execution feedback in the realm of code synthesis, where state-of-the-art LLMs struggle to improve code iteratively compared to independent sampling. We benchmark on competitive programming tasks, where we achieve new start-of-the art results with both small (8B parameters) and large (70B) models while reducing the amount of samples required by an order of magnitude. Our analysis of inference-time behavior demonstrates that our method produces LLMs that effectively leverage automatic feedback over multiple steps.

In this paper, we present a simple but compelling argument, focusing on galaxy science, for preserving the main imagers and operational modes of the Hubble Space Telescope (HST) for as long as is technically feasible. While star-formation started at redshifts z$\gtrsim$10$-$13, when the universe was less than 300$-$500 Myr old, the CSFH did not peak until z$\simeq$1.9, and has steadily declined since that time. Hence, at least half of all stars in the universe formed in the era where HST provides its unique rest-frame UV view of unobscured young, massive stars tracing cosmic star-formation. By rendering a subset of the 556.3 hours of available HST images in 12 filters of the Hubble Ultra Deep Field (HUDF) in an appropriate mix of colors, we illustrate the unique capabilities of HST for galaxy science emphasizing that rest-frame UV$-$optical wavelength range. We then contrast this with the 52.7 publicly available hours of JWST/NIRCam images in 8 filters of the same HUDF area from the JADES project, rendering these at the redder near-IR wavelengths to illustrate the unique capabilities of JWST to detect older stellar populations at higher redshifts, as well as very dusty stellar populations and Active Galactic Nuclei (AGN). HST uniquely probes (unobscured) young, hot, massive stars in galaxies, while JWST reveals more advanced stages of older stellar populations, as well as relatively short-lived phases where galaxies produce and shed a lot of dust from intense star-formation, and the very high redshift universe (z$\gtrsim$10$-$11) not accessible by HST. We conclude that HST and JWST are highly complementary facilities that took decades to build to ensure decades of operation. To maximize return on investment on both HST and JWST, ways will need to be found to operate HST imaging instruments in all relevant modes for as long as possible into the JWST mission., Comment: White paper submitted to NASA/STScI for the HST Senior Review of 2024. The HST+JWST PDF images are too big for astro-ph. Please view their full resolution PDF files on: http://www.asu.edu/clas/hst/www/jwst2024/windhorst_HST_SeniorReview2024_v4.pdf

The fundamental functional summary statistics used for studying spatial point patterns are developed for marked homogeneous and inhomogeneous point processes on the surface of a sphere. These are extended to point processes on the surface of three dimensional convex shapes given the bijective mapping from the shape to the sphere is known. These functional summary statistics are used to test for independence between the marginals of multi-type spatial point processes with methods for sampling the null distribution proposed and discussed. This is illustrated on both simulated data and the RNGC galaxy point pattern, revealing attractive dependencies between different galaxy types.

Dissipative stabilization of cat qubits autonomously corrects for bit flip errors by ensuring that reservoir-engineered two-photon losses dominate over other mechanisms inducing phase flip errors. To describe the latter, we derive an effective master equation for an asymmetrically threaded SQUID based superconducting circuit used to stabilize a dissipative cat qubit. We analyze the dressing of relaxation processes under drives in time-dependent Schrieffer-Wolff perturbation theory for weakly anharmonic bosonic degrees of freedom, and in numerically exact Floquet theory. We find that spurious single-photon decay rates can increase under the action of the parametric pump that generates the required interactions for cat-qubit stabilization. Our analysis feeds into mitigation strategies that can inform current experiments, and the methods presented here can be extended to other circuit implementations., Comment: 24 pages, 13 figures

Reinforcement learning from human feedback (RLHF) has become the leading approach for fine-tuning large language models (LLM). However, RLHF has limitations in multi-task learning (MTL) due to challenges of reward hacking and extreme multi-objective optimization (i.e., trade-off of multiple and/or sometimes conflicting objectives). Applying RLHF for MTL currently requires careful tuning of the weights for reward model and data combinations. This is often done via human intuition and does not generalize. In this work, we introduce a novel post-training paradigm which we called Constrained Generative Policy Optimization (CGPO). The core of CGPO is Mixture of Judges (MoJ) with cost-efficient constrained policy optimization with stratification, which can identify the perfect blend in RLHF in a principled manner. It shows strong empirical results with theoretical guarantees, does not require extensive hyper-parameter tuning, and is plug-and-play in common post-training pipelines. Together, this can detect and mitigate reward hacking behaviors while reaching a pareto-optimal point across an extremely large number of objectives. Our empirical evaluations demonstrate that CGPO significantly outperforms standard RLHF algorithms like PPO and DPO across various tasks including general chat, STEM questions, instruction following, and coding. Specifically, CGPO shows improvements of 7.4% in AlpacaEval-2 (general chat), 12.5% in Arena-Hard (STEM & reasoning), and consistent gains in other domains like math and coding. Notably, PPO, while commonly used, is prone to severe reward hacking in popular coding benchmarks, which CGPO successfully addresses. This breakthrough in RLHF not only tackles reward hacking and extreme multi-objective optimization challenges but also advances the state-of-the-art in aligning general-purpose LLMs for diverse applications., Comment: submitted to conference

The close link between cognitive decline and language has fostered long-standing collaboration between the NLP and medical communities in dementia research. To examine this, we reviewed over 200 papers applying NLP to dementia related efforts, drawing from medical, technological, and NLP-focused literature. We identify key research areas, including dementia detection, linguistic biomarker extraction, caregiver support, and patient assistance, showing that half of all papers focus solely on dementia detection using clinical data. However, many directions remain unexplored: artificially degraded language models, synthetic data, digital twins, and more. We highlight gaps and opportunities around trust, scientific rigor, applicability, and cross-community collaboration, and showcase the diverse datasets encountered throughout our review: recorded, written, structured, spontaneous, synthetic, clinical, social media based, and more. This review aims to inspire more creative approaches to dementia research within the medical and NLP communities.

The generalization error (risk) of a supervised statistical learning algorithm quantifies its prediction ability on previously unseen data. Inspired by exponential tilting, Li et al. (2021) proposed the tilted empirical risk as a non-linear risk metric for machine learning applications such as classification and regression problems. In this work, we examine the generalization error of the tilted empirical risk. In particular, we provide uniform and information-theoretic bounds on the tilted generalization error, defined as the difference between the population risk and the tilted empirical risk, with a convergence rate of $O(1/\sqrt{n})$ where $n$ is the number of training samples. Furthermore, we study the solution to the KL-regularized expected tilted empirical risk minimization problem and derive an upper bound on the expected tilted generalization error with a convergence rate of $O(1/n)$., Comment: 49 pages

Isolated, low-mass galaxies provide the opportunity to assess the impact of reionization on their star formation histories (SFHs) without the ambiguity of environmental processes associated with massive host galaxies. There are very few isolated, low-mass galaxies that are close enough to determine their SFHs from resolved star photometry reaching below the oldest main sequence turnoff. JWST has increased the volume for which this is possible, and here we report on JWST observations of the low-mass, isolated galaxy Leo P. From NIRCam imaging in F090W, F150W, and F277W, we derive a SFH which shows early star formation followed by a pause subsequent to the epoch of reionization which is then later followed by a re-ignition of star formation. This is very similar to the SFHs from previous studies of other dwarf galaxies in the ``transition zone'' between quenched very low-mass galaxies and the more massive galaxies which show no evidence of the impact of reionization on their SFHs; this pattern is rarely produced in simulations of SFHs. The lifetime SFH reveals that Leo P's stellar mass at the epoch of reionization was in the range that is normally associated with being totally quenched. The extended pause in star formation from z~5-1 has important implications for the contribution of low-mass galaxies to the UV photon budget at intermediate redshifts. We also demonstrate that, due to higher sensitivity and angular resolution, observing in two NIRCam short wavelength filters is superior to observing in a combination of a short and a long wavelength filter., Comment: 24 pages, 9 figures, 3 tables

This paper introduces the hypothetical track-length fitting algorithm, a novel method for measuring the kinetic energies of ionizing particles in liquid argon time projection chambers (LArTPCs). The algorithm finds the most probable offset in track length for a track-like object by comparing the measured ionization density as a function of position with a theoretical prediction of the energy loss as a function of the energy, including models of electron recombination and detector response. The algorithm can be used to measure the energies of particles that interact before they stop, such as charged pions that are absorbed by argon nuclei. The algorithm's energy measurement resolutions and fractional biases are presented as functions of particle kinetic energy and number of track hits using samples of stopping secondary charged pions in data collected by the ProtoDUNE-SP detector, and also in a detailed simulation. Additional studies describe impact of the dE/dx model on energy measurement performance. The method described in this paper to characterize the energy measurement performance can be repeated in any LArTPC experiment using stopping secondary charged pions.

Shocks driven by coronal mass ejections (CMEs) are primary drivers of gradual solar energetic particle (SEP) events, posing significant risks to space technology and astronauts. Concurrently, particles accelerated at these shocks may also propagate back to the Sun, potentially generating gamma-ray emissions through pion decay. We incorporated advanced modeling and multi-messenger observations to explore the role of CME-driven shocks in gamma-ray emissions and SEPs. Motivated by Fermi-LAT long-duration solar flares, we used the AWSoM MHD model to investigate the connection between the shocks and the properties of observed gamma-ray emissions. By coupling the AWSoM with iPATH model, we evaluate the impact of shock evolution complexity near the Sun on SEP intensity and spectra. Our result points to the importance of accurate background coronal and solar wind modeling, as well as detailed observations of CME source regions, in advancing our understanding of CME-driven shocks and the dynamics of associated energetic particles., Comment: 11 pages, 6 figures, to appear in the Proceedings of IAU Symposium No. 388 - Solar and Stellar Coronal Mass Ejections

Real-life medical data is often multimodal and incomplete, fueling the growing need for advanced deep learning models capable of integrating them efficiently. The use of diverse modalities, including histopathology slides, MRI, and genetic data, offers unprecedented opportunities to improve prognosis prediction and to unveil new treatment pathways. Contrastive learning, widely used for deriving representations from paired data in multimodal tasks, assumes that different views contain the same task-relevant information and leverages only shared information. This assumption becomes restrictive when handling medical data since each modality also harbors specific knowledge relevant to downstream tasks. We introduce DRIM, a new multimodal method for capturing these shared and unique representations, despite data sparsity. More specifically, given a set of modalities, we aim to encode a representation for each one that can be divided into two components: one encapsulating patient-related information common across modalities and the other, encapsulating modality-specific details. This is achieved by increasing the shared information among different patient modalities while minimizing the overlap between shared and unique components within each modality. Our method outperforms state-of-the-art algorithms on glioma patients survival prediction tasks, while being robust to missing modalities. To promote reproducibility, the code is made publicly available at https://github.com/Lucas-rbnt/DRIM

Training large language models (LLMs) for external tool usage is a rapidly expanding field, with recent research focusing on generating synthetic data to address the shortage of available data. However, the absence of systematic data quality checks poses complications for properly training and testing models. To that end, we propose two approaches for assessing the reliability of data for training LLMs to use external tools. The first approach uses intuitive, human-defined correctness criteria. The second approach uses a model-driven assessment with in-context evaluation. We conduct a thorough evaluation of data quality on two popular benchmarks, followed by an extrinsic evaluation that showcases the impact of data quality on model performance. Our results demonstrate that models trained on high-quality data outperform those trained on unvalidated data, even when trained with a smaller quantity of data. These findings empirically support the significance of assessing and ensuring the reliability of training data for tool-using LLMs.

Quantum Phase Estimation is a crucial component of several front-running quantum algorithms. Improving the efficiency and accuracy of QPE is currently a very active field of research. In this work, we present a hybrid quantum-classical approach that consists of the standard QPE circuit and classical post-processing using curve-fitting, where special attention is given to the latter. We show that our approach achieves high precision with optimal Cram\'er-Rao lower bound performance and is comparable in error resolution with the Variational Quantum Eigensolver and Maximum Likelihood Amplitude Estimation algorithms. Our method could potentially be further extended to the case of estimating multiple phases., Comment: 5 pages, 4 pages of references and appendices, 7 figures

Autonomous systems typically leverage layered control architectures with a combination of discrete and continuous models operating at different timescales. As a result, layered systems form a new class of hybrid systems composed of systems operating on a diverse set of continuous and discrete signals. This paper formalizes the notion of a layered (hierarchical) control architecture through a theory of relations between its layers. This theory enables us to formulate contracts within layered control systems -- these define interfaces between layers and isolate the design of each layer, guaranteeing that composition of contracts at each layer results in a contract capturing the desired system-wide specification. Thus, the proposed theory yields the ability to analyze layered control architectures via a compositional approach.

Quantum query complexity mainly studies the number of queries needed to learn some property of a black box with high probability. A closely related question is how well an algorithm can succeed with this learning task using only a fixed number of queries. In this work, we propose measuring an algorithm's performance using the mutual information between the output and the actual value. A key observation is that if an algorithm is only allowed to make a single query and the goal is to optimize this mutual information, then we obtain a task which is similar to a basic task of quantum communication, where one attempts to maximize the mutual information of the sender and receiver. We make this analogy precise by formally considering the oracle as a separate subsystem, whose state records the unknown oracle identity. The oracle query prepares a state, which is then measured; and the target property of the oracle plays the role of a message that should be deduced from the measurement outcome. Thus we obtain a link between the optimal single-query algorithm and minimization of the extent of quantum correlations between the oracle and the computer subsystems. We also find a lower bound on this mutual information, which is related to quantum coherence. These results extend to multiple-query non-adaptive algorithms. As a result, we gain insight into the task of finding the optimal non-adaptive algorithm that uses at most a fixed number of queries, for any oracle problem., Comment: 19 pages, 1 figure, 5 tables

We demonstrate spectral shaping of entangled photons with a six-channel microring-resonator-based silicon photonic pulse shaper. Through precise calibration of thermal phase shifters in a microresonator-based pulse shaper, we demonstrate line-by-line phase control on a 3~GHz grid for two frequency-bin-entangled qudits, corresponding to Hilbert spaces of up to $6\times 6$ ($3\times 3$) dimensions for shared (independent) signal-idler filters. The pulse shaper's fine spectral resolution enables control of nanosecond-scale temporal features, which are observed by direct coincidence detection of biphoton correlation functions that show excellent agreement with theory. This work marks, to our knowledge, the first demonstration of biphoton pulse shaping using an integrated spectral shaper and holds significant promise for applications in quantum information processing.

High-harmonic generation (HHG) is a nonlinear process in which a material sample is irradiated by intense laser pulses, causing the emission of high harmonics of the incident light. HHG has historically been explained by theories employing a classical electromagnetic field, successfully capturing its spectral and temporal characteristics. However, recent research indicates that quantum-optical effects naturally exist, or can be artificially induced, in HHG. Even though the fundamental equations of motion for quantum electrodynamics (QED) are well-known, a unifying framework for solving them to explore HHG is missing. So far, numerical solutions employed a wide range of basis-sets and untested approximations. Based on methods originally developed for cavity polaritonics, here we formulate a numerically accurate QED model consisting of a single active electron and a single quantized photon mode. Our framework can in principle be extended to higher electronic dimensions and multiple photon modes to be employed in ab initio codes. We employ it as a model of an atom interacting with a photon mode and predict a characteristic minimum structure in the HHG yield vs. phase-squeezing. We find that this phenomenon, which can be used for novel ultrafast quantum spectroscopies, is partially captured by a multi-trajectory Ehrenfest dynamics approach, with the exact minima position sensitive to the level of theory. On the one hand, this motivates using multi-trajectory approaches as an alternative for costly exact calculations. On the other hand, it suggests an inherent limitation of the multi-trajectory formalism, indicating the presence of entanglement. Our work creates a road-map for a universal formalism of QED-HHG that can be employed for benchmarking approximate theories, predicting novel phenomena for advancing quantum applications, and for the measurements of entanglement and entropy., Comment: 10 pages, 7 figures

Ensuring the security of cloud environments is imperative for sustaining organizational growth and operational efficiency. As the ubiquity of cloud services continues to rise, the inevitability of cyber threats underscores the importance of preemptive detection. This paper introduces a pioneering time-based embedding approach for Cloud Services Graph-based Anomaly Detection (CS-GAD), utilizing a Graph Neural Network (GNN) to discern anomalous user behavior during interactions with cloud services. Our method employs a dynamic tripartite graph representation to encapsulate the evolving interactions among cloud services, users, and their activities over time. Leveraging GNN models in each time frame, our approach generates a graph embedding wherein each user is assigned a score based on their historical activity, facilitating the identification of unusual behavior. Results demonstrate a notable reduction in false positive rates (2-9%) compared to prevailing methods, coupled with a commendable true positive rate (100%). The contributions of this work encompass early detection capabilities, a low false positive rate, an innovative tripartite graph representation incorporating action types, the introduction of a new cloud services dataset featuring various user attacks, and an open-source implementation for community collaboration in advancing cloud service security.

Quantifying axon and myelin properties (e.g., axon diameter, myelin thickness, g-ratio) in histology images can provide useful information about microstructural changes caused by neurodegenerative diseases. Automatic tissue segmentation is an important tool for these datasets, as a single stained section can contain up to thousands of axons. Advances in deep learning have made this task quick and reliable with minimal overhead, but a deep learning model trained by one research group will hardly ever be usable by other groups due to differences in their histology training data. This is partly due to subject diversity (different body parts, species, genetics, pathologies) and also to the range of modern microscopy imaging techniques resulting in a wide variability of image features (i.e., contrast, resolution). There is a pressing need to make AI accessible to neuroscience researchers to facilitate and accelerate their workflow, but publicly available models are scarce and poorly maintained. Our approach is to aggregate data from multiple imaging modalities (bright field, electron microscopy, Raman spectroscopy) and species (mouse, rat, rabbit, human), to create an open-source, durable tool for axon and myelin segmentation. Our generalist model makes it easier for researchers to process their data and can be fine-tuned for better performance on specific domains. We study the benefits of different aggregation schemes. This multi-domain segmentation model performs better than single-modality dedicated learners (p=0.03077), generalizes better on out-of-distribution data and is easier to use and maintain. Importantly, we package the segmentation tool into a well-maintained open-source software ecosystem (see https://github.com/axondeepseg/axondeepseg)., Comment: 12 pages, 8 figures

This paper presents a Multi-modal Emotion Recognition (MER) system designed to enhance emotion recognition accuracy in challenging acoustic conditions. Our approach combines a modified and extended Hierarchical Token-semantic Audio Transformer (HTS-AT) for multi-channel audio processing with an R(2+1)D Convolutional Neural Networks (CNN) model for video analysis. We evaluate our proposed method on a reverberated version of the Ryerson audio-visual database of emotional speech and song (RAVDESS) dataset using synthetic and real-world Room Impulse Responsess (RIRs). Our results demonstrate that integrating audio and video modalities yields superior performance compared to uni-modal approaches, especially in challenging acoustic conditions. Moreover, we show that the multimodal (audiovisual) approach that utilizes multiple microphones outperforms its single-microphone counterpart.

We study how the rigidity transition in a triangular lattice changes as a function of anisotropy by preferentially filling bonds on the lattice in one direction. We discover that the onset of rigidity in anisotropic spring networks arises in at least two steps, reminiscent of the two-step melting transition in two dimensional crystals. In particular, our simulations demonstrate that the percolation of stress-supporting bonds happens at different critical volume fractions along different directions. By examining each independent component of the elasticity tensor, we determine universal exponents and develop universal scaling functions to analyze isotropic rigidity percolation as a multicritical point. We expect that these results will be important for elucidating the underlying mechanical phase transitions governing the properties of biological materials ranging from the cytoskeletons of cells to the extracellular networks of tissues such as tendon where the networks are often preferentially aligned., Comment: 13 pages, 17 figures

In this paper, we show that with the ability to jailbreak a GenAI model, attackers can escalate the outcome of attacks against RAG-based GenAI-powered applications in severity and scale. In the first part of the paper, we show that attackers can escalate RAG membership inference attacks and RAG entity extraction attacks to RAG documents extraction attacks, forcing a more severe outcome compared to existing attacks. We evaluate the results obtained from three extraction methods, the influence of the type and the size of five embeddings algorithms employed, the size of the provided context, and the GenAI engine. We show that attackers can extract 80%-99.8% of the data stored in the database used by the RAG of a Q&A chatbot. In the second part of the paper, we show that attackers can escalate the scale of RAG data poisoning attacks from compromising a single GenAI-powered application to compromising the entire GenAI ecosystem, forcing a greater scale of damage. This is done by crafting an adversarial self-replicating prompt that triggers a chain reaction of a computer worm within the ecosystem and forces each affected application to perform a malicious activity and compromise the RAG of additional applications. We evaluate the performance of the worm in creating a chain of confidential data extraction about users within a GenAI ecosystem of GenAI-powered email assistants and analyze how the performance of the worm is affected by the size of the context, the adversarial self-replicating prompt used, the type and size of the embeddings algorithm employed, and the number of hops in the propagation. Finally, we review and analyze guardrails to protect RAG-based inference and discuss the tradeoffs., Comment: for Github, see https://github.com/StavC/UnleashingWorms-ExtractingData

We consider the problem of estimating the direction-of-arrival (DoA) of a desired source located in a known region of interest in the presence of interfering sources and multipath. We propose an approach that precedes the DoA estimation and relies on generating a set of reference steering vectors. The steering vectors' generative model is a free space model, which is beneficial for many DoA estimation algorithms. The set of reference steering vectors is then used to compute a function that maps the received signals from the adverse environment to a reference domain free from interfering sources and multipath. We show theoretically and empirically that the proposed map, which is analogous to domain adaption, improves DoA estimation by mitigating interference and multipath effects. Specifically, we demonstrate a substantial improvement in accuracy when the proposed approach is applied before three commonly used beamformers: the delay-and-sum (DS), the minimum variance distortionless response (MVDR), and the Multiple Signal Classification (MUSIC).

Let $p_1,\ldots,p_n$ be a set of points in the unit square and let $T_1,\ldots,T_n$ be a set of $\delta$-tubes such that $T_j$ passes through $p_j$. We prove a lower bound for the number of incidences between the points and tubes under a natural regularity condition (similar to Frostman regularity). As a consequence, we show that in any configuration of points $p_1,\ldots, p_n \in [0,1]^2$ along with a line $\ell_j$ through each point $p_j$, there exist $j\neq k$ for which $d(p_j, \ell_k) \lesssim n^{-2/3+o(1)}$. It follows from the latter result that any set of $n$ points in the unit square contains three points forming a triangle of area at most $n^{-7/6+o(1)}$. This new upper bound for Heilbronn's triangle problem attains the high-low limit established in our previous work arXiv:2305.18253., Comment: 43 pages, 2 figures

Data from the Decennial Census is published only after applying a disclosure avoidance system (DAS). Data users were shaken by the adoption of differential privacy in the 2020 DAS, a radical departure from past methods. The change raises the question of whether redistricting law permits, forbids, or requires taking account of the effect of disclosure avoidance. Such uncertainty creates legal risks for redistricters, as Alabama argued in a lawsuit seeking to prevent the 2020 DAS's deployment. We consider two redistricting settings in which a data user might be concerned about the impacts of privacy preserving noise: drawing equal population districts and litigating voting rights cases. What discrepancies arise if the user does nothing to account for disclosure avoidance? How might the user adapt her analyses to mitigate those discrepancies? We study these questions by comparing the official 2010 Redistricting Data to the 2010 Demonstration Data -- created using the 2020 DAS -- in an analysis of millions of algorithmically generated state legislative redistricting plans. In both settings, we observe that an analyst may come to incorrect conclusions if they do not account for noise. With minor adaptations, though, the underlying policy goals remain achievable: tweaking selection criteria enables a redistricter to draw balanced plans, and illustrative plans can still be used as evidence of the maximum number of majority-minority districts that are possible in a geography. At least for state legislatures, Alabama's claim that differential privacy ``inhibits a State's right to draw fair lines'' appears unfounded., Comment: 24 pages, 5 figures, 7 tables

We describe a database of 1307 continued fractions with polynomial coefficients, of which more than 1000 are new, both for interesting constants and for transcendental functions, and provide the database inside the \TeX\ source of the paper. Look in particular at the section ``Table of Contents''.

Scaling up quantum computing devices requires solving ever more complex quantum control tasks. Machine learning has been proposed as a promising approach to tackle the resulting challenges. However, experimental implementations are still scarce. In this work, we demonstrate experimentally a neural-network-based preparation of Schr\"odinger cat states in a cavity coupled dispersively to a qubit. We show that it is possible to teach a neural network to output optimized control pulses for a whole family of quantum states. After being trained in simulations, the network takes a description of the target quantum state as input and rapidly produces the pulse shape for the experiment, without any need for time-consuming additional optimization or retraining for different states. Our experimental results demonstrate more generally how deep neural networks and transfer learning can produce efficient simultaneous solutions to a range of quantum control tasks, which will benefit not only state preparation but also parametrized quantum gates., Comment: 20 pages, appendix included

As artificial intelligence (AI) continues to rapidly advance, there is a growing demand to integrate AI capabilities into existing business applications. However, a significant gap exists between the rapid progress in AI and how slowly AI is being embedded into business environments. Deploying well-performing lab models into production settings, especially in on-premise environments, often entails specialized expertise and imposes a heavy burden of model management, creating significant barriers to implementing AI models in real-world applications. KModels leverages proven libraries and platforms (Kubeflow Pipelines, KServe) to streamline AI adoption by supporting both AI developers and consumers. It allows model developers to focus solely on model development and share models as transportable units (Templates), abstracting away complex production deployment concerns. KModels enables AI consumers to eliminate the need for a dedicated data scientist, as the templates encapsulate most data science considerations while providing business-oriented control. This paper presents the architecture of KModels and the key decisions that shape it. We outline KModels' main components as well as its interfaces. Furthermore, we explain how KModels is highly suited for on-premise deployment but can also be used in cloud environments. The efficacy of KModels is demonstrated through the successful deployment of three AI models within an existing Work Order Management system. These models operate in a client's data center and are trained on local data, without data scientist intervention. One model improved the accuracy of Failure Code specification for work orders from 46% to 83%, showcasing the substantial benefit of accessible and localized AI solutions.

Presented here is a transcription of the lecture notes from Professor Allan N. Kaufman's graduate statistical mechanics course at Berkeley from the 1972-1973 academic year. Part 1 addresses equilibrium statistical mechanics with topics: fundamentals, classical fluids and other systems, chemical equilibrium, and long-range interactions. Part 2 addresses non-equilibrium statistical mechanics with topics: fundamentals, Brownian motion, Liouville and Klimontovich equations, Landau equation, Markov processes and Fokker-Planck equation, linear response and transport theory, and an introduction to non-equilibrium quantum statistical mechanics., Comment: 156 pages, lecture notes, a dozen or less figures

Context. On 2022 January 20, the Energetic Particle Detector (EPD) on board Solar Orbiter measured a solar energetic particle (SEP) event showing unusual first arriving particles from the anti-Sun direction. Near-Earth spacecraft separated 17{\deg} in longitude to the west from Solar Orbiter measured classic antisunward-directed fluxes. STEREO-A and MAVEN, separated 18{\deg} to the east and 143{\deg} to the west from Solar Orbiter respectively, also observed the event, suggesting that particles spread over at least 160{\deg} in the heliosphere. Results. Solar Orbiter was embedded in a MC erupting on 16 January from the same active region as the one related to the SEP event on 20 January. The SEP event is related to a M5.5 flare and a fast CME-driven shock of 1433 km/s, which injected particles within and outside the MC. The hard SEP spectra, the presence of a Type II radio burst, and the co-temporal Type III radio bursts being observed from 80 MHz that seems to emanate from the Type II, points to the shock as the relevant accelerator of the particles. Conclusions. The detailed analysis of the SEP event strongly suggest that the energetic particles are injected mainly by a CME-driven shock into and outside of a previous MC present in the heliosphere at the time of the particle onset. The sunward propagating SEPs measured by Solar Orbiter are produced by the injection of particles along the longer (western) leg of the MC still connected to the Sun at the time of the release of the particles. The determined electron propagation path length inside the MC is around 30% longer than the estimated length of the loop leg of the MC itself (based on the graduated cylindrical shell model) consistent with a low number of field line rotations., Comment: 23 pages, 19 figures

To take full advantage of a specific hardware target, performance engineers need to gain control on compilers in order to leverage their domain knowledge about the program and hardware. Yet, modern compilers are poorly controlled, usually by configuring a sequence of coarse-grained monolithic black-box passes, or by means of predefined compiler annotations/pragmas. These can be effective, but often do not let users precisely optimize their varying compute loads. As a consequence, performance engineers have to resort to implementing custom passes for a specific optimization heuristic, requiring compiler engineering expert knowledge. In this paper, we present a technique that provides fine-grained control of general-purpose compilers by introducing the Transform dialect, a controllable IR-based transformation system implemented in MLIR. The Transform dialect empowers performance engineers to optimize their various compute loads by composing and reusing existing - but currently hidden - compiler features without the need to implement new passes or even rebuilding the compiler. We demonstrate in five case studies that the Transform dialect enables precise, safe composition of compiler transformations and allows for straightforward integration with state-of-the-art search methods.

We propose a novel framework for representing neural fields on triangle meshes that is multi-resolution across both spatial and frequency domains. Inspired by the Neural Fourier Filter Bank (NFFB), our architecture decomposes the spatial and frequency domains by associating finer spatial resolution levels with higher frequency bands, while coarser resolutions are mapped to lower frequencies. To achieve geometry-aware spatial decomposition we leverage multiple DiffusionNet components, each associated with a different spatial resolution level. Subsequently, we apply a Fourier feature mapping to encourage finer resolution levels to be associated with higher frequencies. The final signal is composed in a wavelet-inspired manner using a sine-activated MLP, aggregating higher-frequency signals on top of lower-frequency ones. Our architecture attains high accuracy in learning complex neural fields and is robust to discontinuities, exponential scale variations of the target field, and mesh modification. We demonstrate the effectiveness of our approach through its application to diverse neural fields, such as synthetic RGB functions, UV texture coordinates, and vertex normals, illustrating different challenges. To validate our method, we compare its performance against two alternatives, showcasing the advantages of our multi-resolution architecture.

Adiabatic passage in systems of interacting bosons is substantially affected by interactions and inter-particle entanglement. We consider STIRAP-like schemes in Bose-Hubbard chains that exhibit low-dimensional chaos (a 3 site chain), and high-dimensional chaos (more than 3 sites). The dynamics that is generated by a transfer protocol exhibits striking classical and quantum chaos fingerprints that are manifest in the mean-field classical treatment, in the truncated-Wigner semiclassical treatment, and in the full many-body quantum simulations., Comment: 13 pages, 14 figures

How do language models use information provided as context when generating a response? Can we infer whether a particular generated statement is actually grounded in the context, a misinterpretation, or fabricated? To help answer these questions, we introduce the problem of context attribution: pinpointing the parts of the context (if any) that led a model to generate a particular statement. We then present ContextCite, a simple and scalable method for context attribution that can be applied on top of any existing language model. Finally, we showcase the utility of ContextCite through three applications: (1) helping verify generated statements (2) improving response quality by pruning the context and (3) detecting poisoning attacks. We provide code for ContextCite at https://github.com/MadryLab/context-cite.

MICADO is a first light instrument for the Extremely Large Telescope (ELT), set to start operating later this decade. It will provide diffraction limited imaging, astrometry, high contrast imaging, and long slit spectroscopy at near-infrared wavelengths. During the initial phase operations, adaptive optics (AO) correction will be provided by its own natural guide star wavefront sensor. In its final configuration, that AO system will be retained and complemented by the laser guide star multi-conjugate adaptive optics module MORFEO (formerly known as MAORY). Among many other things, MICADO will study exoplanets, distant galaxies and stars, and investigate black holes, such as Sagittarius A* at the centre of the Milky Way. After their final design phase, most components of MICADO have moved on to the manufacturing and assembly phase. Here we summarize the final design of the instrument and provide an overview about its current manufacturing status and the timeline. Some lessons learned from the final design review process will be presented in order to help future instrumentation projects to cope with the challenges arising from the substantial differences between projects for 8-10m class telescopes (e.g. ESO-VLT) and the next generation Extremely Large Telescopes (e.g. ESO-ELT). Finally, the expected performance will be discussed in the context of the current landscape of astronomical observatories and instruments. For instance, MICADO will have similar sensitivity as the James Webb Space Telescope (JWST), but with six times the spatial resolution., Comment: Proceedings of the SPIE, Volume 13096, id. 1309611 11 pp. (2024)

The Virgo Cluster is the nearest substantial cluster of galaxies to the Milky Way and a cornerstone of the extragalactic distance scale. Here, we present JWST/NIRCam observations that simultaneously cover the cores and halos of ten galaxies in and around the Virgo Cluster and are designed to perform simultaneous measurements of the tip of the red giant branch (TRGB) and surface brightness fluctuations (SBF). Seven of the targets are within the Virgo Cluster and where we are able to resolve some of the cluster's substructure, while an additional three provide important constraints on Virgo infall. The seven galaxies within Virgo itself all have SBF measurements from the Advanced Camera for Surveys Virgo Cluster Survey (ACSVCS). After adjusting the ACSVCS measurements for the offset from our new JWST TRGB measurements, we determine a distance to the Virgo Cluster of d $=$ 16.17 $\pm$ 0.25 (stat) $\pm$ 0.47 (sys) Mpc. The work presented here is part of a larger program to develop a Population II distance scale through the TRGB and SBF that is completely independent of the prominent Cepheid + Type Ia supernova ladder. The TRGB distances to the galaxies presented here, when combined with future SBF measurements, will provide a crucial step forward for determining whether or not systematic errors can explain the well-known "Hubble tension'' or if there is significant evidence for cracks in the $\Lambda$CDM model., Comment: submitted to ApJ. comments welcome and appreciated. program name has changed from Paper I's "The Population II Extragalactic Distance Scale" to "The TRGB-SBF Project"

The study of transmission and progression of human papillomavirus (HPV) is crucial for understanding the incidence of cervical cancers, and has been identified as a priority worldwide. The complexity of the disease necessitates a detailed model of HPV transmission and its progression to cancer; to infer properties of the above we require a careful process that can match to imperfect or incomplete observational data. In this paper, we describe the HPVsim simulator to satisfy the former requirement; to satisfy the latter we couple this stochastic simulator to a process of emulation and history matching using the R package hmer. With these tools, we are able to obtain a comprehensive collection of parameter combinations that could give rise to observed cancer data, and explore the implications of the variability of these parameter sets as it relates to future health interventions., Comment: 21 pages, 15 figures; submitted to Epidemics

While Large Language Models (LLMs) are being quickly adapted to many domains, including healthcare, their strengths and pitfalls remain under-explored. In our study, we examine the effects of prompt engineering to guide Large Language Models (LLMs) in delivering parts of a Problem-Solving Therapy (PST) session via text, particularly during the symptom identification and assessment phase for personalized goal setting. We present evaluation results of the models' performances by automatic metrics and experienced medical professionals. We demonstrate that the models' capability to deliver protocolized therapy can be improved with the proper use of prompt engineering methods, albeit with limitations. To our knowledge, this study is among the first to assess the effects of various prompting techniques in enhancing a generalist model's ability to deliver psychotherapy, focusing on overall quality, consistency, and empathy. Exploring LLMs' potential in delivering psychotherapy holds promise with the current shortage of mental health professionals amid significant needs, enhancing the potential utility of AI-based and AI-enhanced care services., Comment: Accepted for AMIA 2024 proceedings