Search

We examine the efficiency of traditional school districts versus charter schools in providing students with teachers who meet their demographic and education needs. Using panel data from the state of Michigan, we estimate the relationship between enrollment of Black, Hispanic, special education, and English learner students and the presence of Black, Hispanic, Special Education, and ESL teachers, and test whether this relationship differs at charter and traditional district-run schools. Because charter schools typically have less market power in hiring than large districts, we compare charter school employment practices to traditional public schools in districts of comparable size. Our results suggest that charter schools are more likely to employ same race teachers for Black students but not Hispanic students, and districts schools are slightly better at providing ESL and SPED teachers. We conclude that charter autonomy does not necessary generate better student-teacher matches, but Michigan charters may occupy a market niche by serving Black students and staffing Black teachers.

In the context of an ongoing national conversation about teacher shortages, we build on prior literature on the efficacy of teacher certification pathways by comparing entry and exit patterns based on age at the time of certification. All trainees who complete a state certification process have invested substantial time and resources into entering teaching. Competing employment opportunities and expectations might vary with age. We use both linear regression and discrete-time hazard models to examine employment and subsequent exit of newly certified teacher trainees in Michigan from 2011 to 2023. We find that while mid-career entrants in their 30s and 40s compose a small share of new certificates, they are more likely to enter a teaching position and no more likely to subsequently exit than counterparts who were certified in their early 20s. Mid-career pathways also contribute to teacher diversity by attracting more Black and male teachers who enter and persist.

In this paper, the issue of averaging data on a manifold is addressed. While the Fr\'echet mean resulting from Riemannian geometry appears ideal, it is unfortunately not always available and often computationally very expensive. To overcome this, R-barycenters have been proposed and successfully applied to Stiefel and Grassmann manifolds. However, R-barycenters still suffer severe limitations as they rely on iterative algorithms and complicated operators. We propose simpler, yet efficient, barycenters that we call RL-barycenters. We show that, in the setting relevant to most applications, our framework yields astonishingly simple barycenters: arithmetic means projected onto the manifold. We apply this approach to the Stiefel and Grassmann manifolds. On simulated data, our approach is competitive with respect to existing averaging methods, while computationally cheaper.

Large-scale Ads recommendation and auction scoring models at Google scale demand immense computational resources. While specialized hardware like TPUs have improved linear algebra computations, bottlenecks persist in large-scale systems. This paper proposes solutions for three critical challenges that must be addressed for efficient end-to-end execution in a widely used production infrastructure: (1) Input Generation and Ingestion Pipeline: Efficiently transforming raw features (e.g., "search query") into numerical inputs and streaming them to TPUs; (2) Large Embedding Tables: Optimizing conversion of sparse features into dense floating-point vectors for neural network consumption; (3) Interruptions and Error Handling: Minimizing resource wastage in large-scale shared datacenters. To tackle these challenges, we propose a shared input generation technique to reduce computational load of input generation by amortizing costs across many models. Furthermore, we propose partitioning, pipelining, and RPC (Remote Procedure Call) coalescing software techniques to optimize embedding operations. To maintain efficiency at scale, we describe novel preemption notice and training hold mechanisms that minimize resource wastage, and ensure prompt error resolution. These techniques have demonstrated significant improvement in Google production, achieving a 116% performance boost and an 18% reduction in training costs across representative models., Comment: 13 pages, 7 figures

Understanding how large language models (LLMs) grasp the historical context of concepts and their semantic evolution is essential in advancing artificial intelligence and linguistic studies. This study aims to evaluate the capabilities of various LLMs in capturing temporal dynamics of meaning, specifically how they interpret terms across different time periods. We analyze a diverse set of terms from multiple domains, using tailored prompts and measuring responses through both objective metrics (e.g., perplexity and word count) and subjective human expert evaluations. Our comparative analysis includes prominent models like ChatGPT, GPT-4, Claude, Bard, Gemini, and Llama. Findings reveal marked differences in each model's handling of historical context and semantic shifts, highlighting both strengths and limitations in temporal semantic understanding. These insights offer a foundation for refining LLMs to better address the evolving nature of language, with implications for historical text analysis, AI design, and applications in digital humanities.

Electrocardiograms (ECGs) are essential for diagnosing cardiac pathologies, yet traditional paper-based ECG storage poses significant challenges for automated analysis. This study introduces ECGtizer, an open-source, fully automated tool designed to digitize paper ECGs and recover signals lost during storage. ECGtizer facilitates automated analyses using modern AI methods. It employs automated lead detection, three pixel-based signal extraction algorithms, and a deep learning-based signal reconstruction module. We evaluated ECGtizer on two datasets: a real-life cohort from the COVID-19 pandemic (JOCOVID) and a publicly available dataset (PTB-XL). Performance was compared with two existing methods: the fully automated ECGminer and the semi-automated PaperECG, which requires human intervention. ECGtizer's performance was assessed in terms of signal recovery and the fidelity of clinically relevant feature measurement. Additionally, we tested these tools on a third dataset (GENEREPOL) for downstream AI tasks. Results show that ECGtizer outperforms existing tools, with its ECGtizerFrag algorithm delivering superior signal recovery. While PaperECG demonstrated better outcomes than ECGminer, it required human input. ECGtizer enhances the usability of historical ECG data and supports advanced AI-based diagnostic methods, making it a valuable addition to the field of AI in ECG analysis.

High-Performance Computing (HPC) systems are among the most energy-intensive scientific facilities, with electric power consumption reaching and often exceeding 20 megawatts per installation. Unlike other major scientific infrastructures such as particle accelerators or high-intensity light sources, which are few around the world, the number and size of supercomputers are continuously increasing. Even if every new system generation is more energy efficient than the previous one, the overall growth in size of the HPC infrastructure, driven by a rising demand for computational capacity across all scientific disciplines, and especially by artificial intelligence workloads (AI), rapidly drives up the energy demand. This challenge is particularly significant for HPC centers in Germany, where high electricity costs, stringent national energy policies, and a strong commitment to environmental sustainability are key factors. This paper describes various state-of-the-art strategies and innovations employed to enhance the energy efficiency of HPC systems within the national context. Case studies from leading German HPC facilities illustrate the implementation of novel heterogeneous hardware architectures, advanced monitoring infrastructures, high-temperature cooling solutions, energy-aware scheduling, and dynamic power management, among other optimizations. By reviewing best practices and ongoing research, this paper aims to share valuable insight with the global HPC community, motivating the pursuit of more sustainable and energy-efficient HPC operations., Comment: 30 pages, 3 figures, 4 tables

We consider an inverse boundary value problem for the biharmonic operator with the first order perturbation in a bounded domain of dimension three or higher. Assuming that the first and the zeroth order perturbations are known in a neighborhood of the boundary, we establish log-type stability estimates for these perturbations from a partial Dirichlet-to-Neumann map. Specifically, measurements are taken only on an arbitrarily small open subsets of the boundary.

Significant progress has been made in advancing large multimodal conversational models (LMMs), capitalizing on vast repositories of image-text data available online. Despite this progress, these models often encounter substantial domain gaps, hindering their ability to engage in complex conversations across new domains. Recent efforts have aimed to mitigate this issue, albeit relying on domain-specific image-text data to curate instruction-tuning data. However, many domains, such as agriculture, lack such vision-language data. In this work, we propose an approach to construct instruction-tuning data that harnesses vision-only data for the agriculture domain. We utilize diverse agricultural datasets spanning multiple domains, curate class-specific information, and employ large language models (LLMs) to construct an expert-tuning set, resulting in a 70k expert-tuning dataset called AgroInstruct. Subsequently, we expert-tuned and created AgroGPT, an efficient LMM that can hold complex agriculture-related conversations and provide useful insights. We also develop AgroEvals for evaluation and compare {AgroGPT's} performance with large open and closed-source models. {AgroGPT} excels at identifying fine-grained agricultural concepts, can act as an agriculture expert, and provides helpful information for multimodal agriculture questions. The code, datasets, and models are available at https://github.com/awaisrauf/agroGPT., Comment: Accepted at WACV, 2025

Mangroves are dynamic coastal ecosystems that are crucial to environmental health, economic stability, and climate resilience. The monitoring and preservation of mangroves are of global importance, with remote sensing technologies playing a pivotal role in these efforts. The integration of cutting-edge artificial intelligence with satellite data opens new avenues for ecological monitoring, potentially revolutionizing conservation strategies at a time when the protection of natural resources is more crucial than ever. The objective of this work is to provide a comprehensive evaluation of recent deep-learning models on the task of mangrove segmentation. We first introduce and make available a novel open-source dataset, MagSet-2, incorporating mangrove annotations from the Global Mangrove Watch and satellite images from Sentinel-2, from mangrove positions all over the world. We then benchmark three architectural groups, namely convolutional, transformer, and mamba models, using the created dataset. The experimental outcomes further validate the deep learning community's interest in the Mamba model, which surpasses other architectures in all metrics., Comment: 12 pages, accepted to the MACLEAN workshop of ECML/PKDD 2024

Large Language Models (LLMs) are rapidly becoming commodity components of larger software systems. This poses natural security and privacy problems: poisoned data retrieved from one component can change the model's behavior and compromise the entire system, including coercing the model to spread confidential data to untrusted components. One promising approach is to tackle this problem at the system level via dynamic information flow (aka taint) tracking. Unfortunately, the traditional approach of propagating the most restrictive input label to the output is too conservative for applications where LLMs operate on inputs retrieved from diverse sources. In this paper, we propose a novel, more permissive approach to propagate information flow labels through LLM queries. The key idea behind our approach is to propagate only the labels of the samples that were influential in generating the model output and to eliminate the labels of unnecessary input. We implement and investigate the effectiveness of two variations of this approach, based on (i) prompt-based retrieval augmentation, and (ii) a $k$-nearest-neighbors language model. We compare these with the baseline of an introspection-based influence estimator that directly asks the language model to predict the output label. The results obtained highlight the superiority of our prompt-based label propagator, which improves the label in more than 85% of the cases in an LLM agent setting. These findings underscore the practicality of permissive label propagation for retrieval augmentation., Comment: 16 pages, 11 figures

Vision-language models (VLMs) are essential for contextual understanding of both visual and textual information. However, their vulnerability to adversarially manipulated inputs presents significant risks, leading to compromised outputs and raising concerns about the reliability in VLM-integrated applications. Detecting these malicious prompts is thus crucial for maintaining trust in VLM generations. A major challenge in developing a safeguarding prompt classifier is the lack of a large amount of labeled benign and malicious data. To address the issue, we introduce VLMGuard, a novel learning framework that leverages the unlabeled user prompts in the wild for malicious prompt detection. These unlabeled prompts, which naturally arise when VLMs are deployed in the open world, consist of both benign and malicious information. To harness the unlabeled data, we present an automated maliciousness estimation score for distinguishing between benign and malicious samples within this unlabeled mixture, thereby enabling the training of a binary prompt classifier on top. Notably, our framework does not require extra human annotations, offering strong flexibility and practicality for real-world applications. Extensive experiment shows VLMGuard achieves superior detection results, significantly outperforming state-of-the-art methods. Disclaimer: This paper may contain offensive examples; reader discretion is advised., Comment: arXiv admin note: text overlap with arXiv:2409.17504

Benchmarks are essential in the design of modern HPC installations, as they define key aspects of system components. Beyond synthetic workloads, it is crucial to include real applications that represent user requirements into benchmark suites, to guarantee high usability and widespread adoption of a new system. Given the significant investments in leadership-class supercomputers of the exascale era, this is even more important and necessitates alignment with a vision of Open Science and reproducibility. In this work, we present the JUPITER Benchmark Suite, which incorporates 16 applications from various domains. It was designed for and used in the procurement of JUPITER, the first European exascale supercomputer. We identify requirements and challenges and outline the project and software infrastructure setup. We provide descriptions and scalability studies of selected applications and a set of key takeaways. The JUPITER Benchmark Suite is released as open source software with this work at https://github.com/FZJ-JSC/jubench., Comment: To be published in Proceedings of The International Conference for High Performance Computing Networking, Storage, and Analysis (SC '24) (2024)

With increasingly larger and more complex telecommunication networks, there is a need for improved monitoring and reliability. Requirements increase further when working with mission-critical systems requiring stable operations to meet precise design and client requirements while maintaining high availability. This paper proposes a novel methodology for developing a machine learning model that can assist in maintaining availability (through anomaly detection) for client-server communications in mission-critical systems. To that end, we validate our methodology for training models based on data classified according to client performance. The proposed methodology evaluates the use of machine learning to perform anomaly detection of a single virtualized server loaded with simulated network traffic (using SIPp) with media calls. The collected data for the models are classified based on the round trip time performance experienced on the client side to determine if the trained models can detect anomalous client side performance only using key performance indicators available on the server. We compared the performance of seven different machine learning models by testing different trained and untrained test stressor scenarios. In the comparison, five models achieved an F1-score above 0.99 for the trained test scenarios. Random Forest was the only model able to attain an F1-score above 0.9 for all untrained test scenarios with the lowest being 0.980. The results suggest that it is possible to generate accurate anomaly detection to evaluate degraded client-side performance.

In this paper, we study a well known two-step anaerobic digestion model in a configuration of two chemostats in series. This model is an eight-dimensional system of ordinary differential equations. Since the reaction system has a cascade structure, we show that the eight-order model can be reduced to a four-dimensional one. Using general growth rates, we provide an in-depth mathematical analysis of the asymptotic behavior of the system. First, we determine all the steady states of the model where there can be more than fifteen equilibria with a non-monotonic growth rate. Then, the necessary and sufficient conditions of existence and local stability of all steady states are established according to the operating parameters: the dilution rate, the input concentrations of the two nutrients, and the distribution of the total process volume considered. The operating diagrams are then analyzed theoretically to describe the asymptotic behavior of the process according to the four control parameters. There can be seventy regions with rich behavior where the system may exhibit bistability or tristability with the coexistence of both microbial species in the two bioreactors.

Since Galileo and (more recently) D'Arcy Thompson, it has been understood that physical processes and constraints influence biological structures and their resulting functions. However these cross-discpline connections -- and their importance to growing scientific discplines such as biophysics -- are rarely tought in introductory physics courses. Here we examine how the laws of physics shape Darwinism evolution through the surface area to volume ratio, an important geometric measure of a structure. We develop conceptual cartoon clicker questions to enhance students' understanding of these interdisciplinary concepts. By connecting abstract physical laws with biological (and technological) applications, our approach aims to help students appreciate the deep connections between physical and biological sciences, thereby enriching the learning experience in introductory physics courses., Comment: 31 page, 17 figures

Safe, all-solid-state lithium metal batteries enable high energy density applications, but suffer from instabilities during operation that lead to rough interfaces between the metal and electrolyte and subsequently cause void formation and dendrite growth that degrades performance and safety. Inspired by the morphogenetic control of thin lamina such as tree leaves that robustly grow into flat shapes -- we propose a range of approaches to control lithium metal stripping and plating. To guide discovery of materials that will implement these feedback mechanisms, we develop a reduced order model that captures couplings between mechanics, interface growth, temperature, and electrochemical variables. We find that long-range feedback is required to achieve true interface stability, while approaches based on local feedback always eventually grow into rough interfaces. All together, our study provides the beginning of a practical framework for analyzing and designing stable electrochemical interfaces in terms of the mechanical properties and the physical chemistry that underlie their dynamics.

This is the third in a series of annual reports about teacher shortages in Michigan that the state legislature requested in December 2020 (2020 PA 316). Although the state data on this topic is limited, these analyses still help to paint a picture of teacher shortages across Michigan, assist policymakers to target policies and programs in ways that can best support the state and local communities in growing their teacher workforces, and highlight ways that new or better data may provide a deeper understanding of local and statewide teacher shortages. In addition to updating the analyses from EPIC's first comprehensive report, we continue to adjust and expand on our past analyses based on the results from prior reports and any additions or improvements to the data available each year.

The increasing cost of training machine learning (ML) models has led to the inclusion of new parties to the training pipeline, such as users who contribute training data and companies that provide computing resources. This involvement of such new parties in the ML training process has introduced new attack surfaces for an adversary to exploit. A recent attack in this domain is the model hijacking attack, whereby an adversary hijacks a victim model to implement their own -- possibly malicious -- hijacking tasks. However, the scope of the model hijacking attack is so far limited to the homogeneous-modality tasks. In this paper, we transform the model hijacking attack into a more general multimodal setting, where the hijacking and original tasks are performed on data of different modalities. Specifically, we focus on the setting where an adversary implements a natural language processing (NLP) hijacking task into an image classification model. To mount the attack, we propose a novel encoder-decoder based framework, namely the Blender, which relies on advanced image and language models. Experimental results show that our modal hijacking attack achieves strong performances in different settings. For instance, our attack achieves 94%, 94%, and 95% attack success rate when using the Sogou news dataset to hijack STL10, CIFAR-10, and MNIST classifiers.

Recently, autonomous agents built on large language models (LLMs) have experienced significant development and are being deployed in real-world applications. These agents can extend the base LLM's capabilities in multiple ways. For example, a well-built agent using GPT-3.5-Turbo as its core can outperform the more advanced GPT-4 model by leveraging external components. More importantly, the usage of tools enables these systems to perform actions in the real world, moving from merely generating text to actively interacting with their environment. Given the agents' practical applications and their ability to execute consequential actions, it is crucial to assess potential vulnerabilities. Such autonomous systems can cause more severe damage than a standalone language model if compromised. While some existing research has explored harmful actions by LLM agents, our study approaches the vulnerability from a different perspective. We introduce a new type of attack that causes malfunctions by misleading the agent into executing repetitive or irrelevant actions. We conduct comprehensive evaluations using various attack methods, surfaces, and properties to pinpoint areas of susceptibility. Our experiments reveal that these attacks can induce failure rates exceeding 80\% in multiple scenarios. Through attacks on implemented and deployable agents in multi-agent scenarios, we accentuate the realistic risks associated with these vulnerabilities. To mitigate such attacks, we propose self-examination detection methods. However, our findings indicate these attacks are difficult to detect effectively using LLMs alone, highlighting the substantial risks associated with this vulnerability.

Plasma in the terrestrial magnetosheath is characterised by very weak particle-particle collisions, so kinetic microinstabilities are thought to be responsible for regulating the thermodynamics of the plasma. By exciting electromagnetic waves, these instabilities redistribute free energy in velocity space, moulding the velocity distribution function (VDF) into a lower energy state. In the high-beta magnetosheath, relatively small perturbations to the VDF can easily excite instabilities compared to in the low-beta inner heliosphere. Since magnetic fields cannot do work on the particles, electric fields mediate energy exchange between the electromagnetic field and the bulk fluid properties of the plasma. We investigate signatures of non-ideal energy conversion associated with turbulent fluctuations in the context of electron and ion temperature anisotropy-beta instabilities, utilising over 24 hours of data spread over 163 distinct intervals of in situ magnetosheath observations from Magnetospheric Multiscale (MMS). We find that average energy conversion into fluid flow is enhanced along instability boundaries, suggesting that turbulence is playing a role in how free energy is redistributed in the plasma. The work enables a quantification of the energetics which are associated with the role of kinetic microinstabilities in regulating collisionless plasma thermodynamics. This work provides insight into the open question of how specific plasma processes couple into the turbulent dynamics and ultimately lead to energy dissipation and particle energisation in collisionless plasmas., Comment: 10 pages, 3 figures, 1 table. Submitted for peer review at Astrophysical Journal Letters (ApJL), July 2024

Amid growing concerns over the ease of theft and misuse of Large Language Models (LLMs), the need for fingerprinting models has increased. Fingerprinting, in this context, means that the model owner can link a given model to their original version, thereby identifying if their model is being misused or has been completely stolen. In this paper, we first define a set five properties a successful fingerprint should satisfy; namely, the fingerprint should be Transparent, Efficient, Persistent, Robust, and Unforgeable. Next, we propose Chain & Hash, a new, simple fingerprinting approach that implements a fingerprint with a cryptographic flavor, achieving all these properties. Chain & Hash involves generating a set of questions (the fingerprints) along with a set of potential answers. These elements are hashed together using a secure hashing technique to select the value for each question, hence providing an unforgeability property-preventing adversaries from claiming false ownership. We evaluate the Chain & Hash technique on multiple models and demonstrate its robustness against benign transformations, such as fine-tuning on different datasets, and adversarial attempts to erase the fingerprint. Finally, our experiments demonstrate the efficiency of implementing Chain & Hash and its utility, where fingerprinted models achieve almost the same performance as non-fingerprinted ones across different benchmarks.

Large Language Models (LLMs) are gaining momentum in software development with prompt-driven programming enabling developers to create code from natural language (NL) instructions. However, studies have questioned their ability to produce secure code and, thereby, the quality of prompt-generated software. Alongside, various prompting techniques that carefully tailor prompts have emerged to elicit optimal responses from LLMs. Still, the interplay between such prompting strategies and secure code generation remains under-explored and calls for further investigations. OBJECTIVE: In this study, we investigate the impact of different prompting techniques on the security of code generated from NL instructions by LLMs. METHOD: First we perform a systematic literature review to identify the existing prompting techniques that can be used for code generation tasks. A subset of these techniques are evaluated on GPT-3, GPT-3.5, and GPT-4 models for secure code generation. For this, we used an existing dataset consisting of 150 NL security-relevant code-generation prompts. RESULTS: Our work (i) classifies potential prompting techniques for code generation (ii) adapts and evaluates a subset of the identified techniques for secure code generation tasks and (iii) observes a reduction in security weaknesses across the tested LLMs, especially after using an existing technique called Recursive Criticism and Improvement (RCI), contributing valuable insights to the ongoing discourse on LLM-generated code security., Comment: This work was partially supported by the EU-funded project Sec4AI4Sec: Cybersecurity for AI-Augmented Systems (grant no. 101120393)

This paper presents a study on autonomous robot navigation, focusing on three key behaviors: Odometry, Target Tracking, and Obstacle Avoidance. Each behavior is described in detail, along with experimental setups for simulated and real-world environments. Odometry utilizes wheel encoder data for precise navigation along predefined paths, validated through experiments with a Pioneer robot. Target Tracking employs vision-based techniques for pursuing designated targets while avoiding obstacles, demonstrated on the same platform. Obstacle Avoidance utilizes ultrasonic sensors to navigate cluttered environments safely, validated in both simulated and real-world scenarios. Additionally, the paper extends the project to include an Elegoo robot car, leveraging its features for enhanced experimentation. Through advanced algorithms and experimental validations, this study provides insights into developing robust navigation systems for autonomous robots.

Open-source large language models (LLMs) have become increasingly popular among both the general public and industry, as they can be customized, fine-tuned, and freely used. However, some open-source LLMs require approval before usage, which has led to third parties publishing their own easily accessible versions. Similarly, third parties have been publishing fine-tuned or quantized variants of these LLMs. These versions are particularly appealing to users because of their ease of access and reduced computational resource demands. This trend has increased the risk of training time attacks, compromising the integrity and security of LLMs. In this work, we present a new training time attack, SOS, which is designed to be low in computational demand and does not require clean data or modification of the model weights, thereby maintaining the model's utility intact. The attack addresses security issues in various scenarios, including the backdoor attack, jailbreak attack, and prompt stealing attack. Our experimental findings demonstrate that the proposed attack is effective across all evaluated targets. Furthermore, we present the other side of our SOS technique, namely the copyright token -- a novel technique that enables users to mark their copyrighted content and prevent models from using it.

Generative Flow Networks (GFlowNets) have emerged as an innovative learning paradigm designed to address the challenge of sampling from an unnormalized probability distribution, called the reward function. This framework learns a policy on a constructed graph, which enables sampling from an approximation of the target probability distribution through successive steps of sampling from the learned policy. To achieve this, GFlowNets can be trained with various objectives, each of which can lead to the model s ultimate goal. The aspirational strength of GFlowNets lies in their potential to discern intricate patterns within the reward function and their capacity to generalize effectively to novel, unseen parts of the reward function. This paper attempts to formalize generalization in the context of GFlowNets, to link generalization with stability, and also to design experiments that assess the capacity of these models to uncover unseen parts of the reward function. The experiments will focus on length generalization meaning generalization to states that can be constructed only by longer trajectories than those seen in training.

Data limitation is a significant challenge in applying deep learning to medical images. Recently, the diffusion probabilistic model (DPM) has shown the potential to generate high-quality images by converting Gaussian random noise into realistic images. In this paper, we apply the DPM to augment the deep ultraviolet fluorescence (DUV) image dataset with an aim to improve breast cancer classification for intraoperative margin assessment. For classification, we divide the whole surface DUV image into small patches and extract convolutional features for each patch by utilizing the pre-trained ResNet. Then, we feed them into an XGBoost classifier for patch-level decisions and then fuse them with a regional importance map computed by Grad-CAM++ for whole surface-level prediction. Our experimental results show that augmenting the training dataset with the DPM significantly improves breast cancer detection performance in DUV images, increasing accuracy from 93% to 97%, compared to using Affine transformations and ProGAN., Comment: IEEE International Symposium on Biomedical Imaging 2024

Preference optimization methods have been successfully applied to improve not only the alignment of large language models (LLMs) with human values, but also specific natural language tasks such as summarization and stylistic continuations. This paper proposes using preference optimization methods on Chain-of-Thought steps in order to improve the reasoning performances of language models. While the chosen answers are obtained from datasets that include reasoning traces, we propose two complementary schemes for generating rejected answers: digit corruption, and weak LLM prompting. Our approach leads to increased accuracy on the GSM8K, AQuA-RAT, and ARC benchmarks for Falcon2-11B and Mistral-7B. For example, the approach can lead to up to a relative 8.47% increase in accuracy on the GSM8K benchmark without any extra annotations. This work suggests that spending resources on creating more datasets of reasoning traces would further boost LLM performances on informal reasoning tasks.

Current research in breast cancer Magnetic Resonance Imaging (MRI), especially with Artificial Intelligence (AI), faces challenges due to the lack of expert segmentations. To address this, we introduce the MAMA-MIA dataset, comprising 1506 multi-center dynamic contrast-enhanced MRI cases with expert segmentations of primary tumors and non-mass enhancement areas. These cases were sourced from four publicly available collections in The Cancer Imaging Archive (TCIA). Initially, we trained a deep learning model to automatically segment the cases, generating preliminary segmentations that significantly reduced expert segmentation time. Sixteen experts, averaging 9 years of experience in breast cancer, then corrected these segmentations, resulting in the final expert segmentations. Additionally, two radiologists conducted a visual inspection of the automatic segmentations to support future quality control studies. Alongside the expert segmentations, we provide 49 harmonized demographic and clinical variables and the pretrained weights of the well-known nnUNet architecture trained using the DCE-MRI full-images and expert segmentations. This dataset aims to accelerate the development and benchmarking of deep learning models and foster innovation in breast cancer diagnostics and treatment planning., Comment: 15 paes, 7 figures, 3 tables

As the pace of AI technology continues to accelerate, more tools have become available to researchers to solve longstanding problems, Hybrid approaches available today continue to push the computational limits of efficiency and precision. One of such problems is the inverse kinematics of redundant systems. This paper explores the complexities of a 7 degree of freedom manipulator and explores 13 optimization techniques to solve it. Additionally, a novel approach is proposed to contribute to the field of algorithmic research. This was found to be over 200 times faster than the well-known traditional Particle Swarm Optimization technique. This new method may serve as a new field of search that combines the explorative capabilities of Machine Learning with the exploitative capabilities of numerical methods.

In the intricate field of medical diagnostics, capturing the subtle manifestations of diseases remains a challenge. Traditional methods, often binary in nature, may not encapsulate the nuanced variances that exist in real-world clinical scenarios. This paper introduces a novel approach by leveraging Fuzzy Logic Rules to derive disease classes based on expert domain knowledge from a medical practitioner. By recognizing that diseases do not always fit into neat categories, and that expert knowledge can guide the fuzzification of these boundaries, our methodology offers a more sophisticated and nuanced diagnostic tool. Using a dataset procured from a prominent hospital, containing detailed patient blood count records, we harness Fuzzy Logic Rules, a computational technique celebrated for its ability to handle ambiguity. This approach, moving through stages of fuzzification, rule application, inference, and ultimately defuzzification, produces refined diagnostic predictions. When combined with the Random Forest classifier, the system adeptly predicts hematological conditions using Complete Blood Count (CBC) parameters. Preliminary results showcase high accuracy levels, underscoring the advantages of integrating fuzzy logic into the diagnostic process. When juxtaposed with traditional diagnostic techniques, it becomes evident that Fuzzy Logic, especially when guided by medical expertise, offers significant advancements in the realm of hematological diagnostics. This paper not only paves the path for enhanced patient care but also beckons a deeper dive into the potentialities of fuzzy logic in various medical diagnostic applications.

Large language model systems face important security risks from maliciously crafted messages that aim to overwrite the system's original instructions or leak private data. To study this problem, we organized a capture-the-flag competition at IEEE SaTML 2024, where the flag is a secret string in the LLM system prompt. The competition was organized in two phases. In the first phase, teams developed defenses to prevent the model from leaking the secret. During the second phase, teams were challenged to extract the secrets hidden for defenses proposed by the other teams. This report summarizes the main insights from the competition. Notably, we found that all defenses were bypassed at least once, highlighting the difficulty of designing a successful defense and the necessity for additional research to protect LLM systems. To foster future research in this direction, we compiled a dataset with over 137k multi-turn attack chats and open-sourced the platform.