Maya Pavlova

Aaron Grattafiori, Abhimanyu Dubey, Abhinav Jauhri et al. · allen-ai, berkeley

Modern artificial intelligence (AI) systems are powered by foundation models. This paper presents a new set of foundation models, called Llama 3. It is a herd of language models that natively support multilinguality, coding, reasoning, and tool usage. Our largest model is a dense Transformer with 405B parameters and a context window of up to 128K tokens. This paper presents an extensive empirical evaluation of Llama 3. We find that Llama 3 delivers comparable quality to leading language models such as GPT-4 on a plethora of tasks. We publicly release Llama 3, including pre-trained and post-trained versions of the 405B parameter language model and our Llama Guard 3 model for input and output safety. The paper also presents the results of experiments in which we integrate image, video, and speech capabilities into Llama 3 via a compositional approach. We observe this approach performs competitively with the state-of-the-art on image, video, and speech recognition tasks. The resulting models are not yet being broadly released as they are still under development.

Hossein Aboutalebi, Maya Pavlova, Mohammad Javad Shafiee et al.

Since the World Health Organization declared COVID-19 a pandemic in 2020, the global community has faced ongoing challenges in controlling and mitigating the transmission of the SARS-CoV-2 virus, as well as its evolving subvariants and recombinants. A significant challenge during the pandemic has not only been the accurate detection of positive cases but also the efficient prediction of risks associated with complications and patient survival probabilities. These tasks entail considerable clinical resource allocation and attention.In this study, we introduce COVID-Net Biochem, a versatile and explainable framework for constructing machine learning models. We apply this framework to predict COVID-19 patient survival and the likelihood of developing Acute Kidney Injury during hospitalization, utilizing clinical and biochemical data in a transparent, systematic approach. The proposed approach advances machine learning model design by seamlessly integrating domain expertise with explainability tools, enabling model decisions to be based on key biomarkers. This fosters a more transparent and interpretable decision-making process made by machines specifically for medical applications.

Maya Pavlova, Tia Tuinstra, Hossein Aboutalebi et al.

After more than two years since the beginning of the COVID-19 pandemic, the pressure of this crisis continues to devastate globally. The use of chest X-ray (CXR) imaging as a complementary screening strategy to RT-PCR testing is not only prevailing but has greatly increased due to its routine clinical use for respiratory complaints. Thus far, many visual perception models have been proposed for COVID-19 screening based on CXR imaging. Nevertheless, the accuracy and the generalization capacity of these models are very much dependent on the diversity and the size of the dataset they were trained on. Motivated by this, we introduce COVIDx CXR-3, a large-scale benchmark dataset of CXR images for supporting COVID-19 computer vision research. COVIDx CXR-3 is composed of 30,386 CXR images from a multinational cohort of 17,026 patients from at least 51 countries, making it, to the best of our knowledge, the most extensive, most diverse COVID-19 CXR dataset in open access form. Here, we provide comprehensive details on the various aspects of the proposed dataset including patient demographics, imaging views, and infection types. The hope is that COVIDx CXR-3 can assist scientists in advancing machine learning research against both the COVID-19 pandemic and related diseases.

Joanna Bitton, Maya Pavlova, Ivan Evtimov

Text-based adversarial attacks are becoming more commonplace and accessible to general internet users. As these attacks proliferate, the need to address the gap in model robustness becomes imminent. While retraining on adversarial data may increase performance, there remains an additional class of character-level attacks on which these models falter. Additionally, the process to retrain a model is time and resource intensive, creating a need for a lightweight, reusable defense. In this work, we propose the Adversarial Text Normalizer, a novel method that restores baseline performance on attacked content with low computational overhead. We evaluate the efficacy of the normalizer on two problem areas prone to adversarial attacks, i.e. Hate Speech and Natural Language Inference. We find that text normalization provides a task-agnostic defense against character-level attacks that can be implemented supplementary to adversarial retraining solutions, which are more suited for semantic alterations.

Alexander MacLean, Saad Abbasi, Ashkan Ebadi et al.

The Coronavirus Disease 2019 (COVID-19) pandemic has impacted many aspects of life globally, and a critical factor in mitigating its effects is screening individuals for infections, thereby allowing for both proper treatment for those individuals as well as action to be taken to prevent further spread of the virus. Point-of-care ultrasound (POCUS) imaging has been proposed as a screening tool as it is a much cheaper and easier to apply imaging modality than others that are traditionally used for pulmonary examinations, namely chest x-ray and computed tomography. Given the scarcity of expert radiologists for interpreting POCUS examinations in many highly affected regions around the world, low-cost deep learning-driven clinical decision support solutions can have a large impact during the on-going pandemic. Motivated by this, we introduce COVID-Net US, a highly efficient, self-attention deep convolutional neural network design tailored for COVID-19 screening from lung POCUS images. Experimental results show that the proposed COVID-Net US can achieve an AUC of over 0.98 while achieving 353X lower architectural complexity, 62X lower computational complexity, and 14.3X faster inference times on a Raspberry Pi. Clinical validation was also conducted, where select cases were reviewed and reported on by a practicing clinician (20 years of clinical practice) specializing in intensive care (ICU) and 15 years of expertise in POCUS interpretation. To advocate affordable healthcare and artificial intelligence for resource-constrained environments, we have made COVID-Net US open source and publicly available as part of the COVID-Net open source initiative.

Maya Pavlova, Naomi Terhljan, Audrey G. Chung et al.

As the COVID-19 pandemic continues to devastate globally, the use of chest X-ray (CXR) imaging as a complimentary screening strategy to RT-PCR testing continues to grow given its routine clinical use for respiratory complaint. As part of the COVID-Net open source initiative, we introduce COVID-Net CXR-2, an enhanced deep convolutional neural network design for COVID-19 detection from CXR images built using a greater quantity and diversity of patients than the original COVID-Net. To facilitate this, we also introduce a new benchmark dataset composed of 19,203 CXR images from a multinational cohort of 16,656 patients from at least 51 countries, making it the largest, most diverse COVID-19 CXR dataset in open access form. The COVID-Net CXR-2 network achieves sensitivity and positive predictive value of 95.5%/97.0%, respectively, and was audited in a transparent and responsible manner. Explainability-driven performance validation was used during auditing to gain deeper insights in its decision-making behaviour and to ensure clinically relevant factors are leveraged for improving trust in its usage. Radiologist validation was also conducted, where select cases were reviewed and reported on by two board-certified radiologists with over 10 and 19 years of experience, respectively, and showed that the critical factors leveraged by COVID-Net CXR-2 are consistent with radiologist interpretations. While not a production-ready solution, we hope the open-source, open-access release of COVID-Net CXR-2 and the respective CXR benchmark dataset will encourage researchers, clinical scientists, and citizen scientists to accelerate advancements and innovations in the fight against the pandemic.

Hossein Aboutalebi, Maya Pavlova, Mohammad Javad Shafiee et al.

The world is still struggling in controlling and containing the spread of the COVID-19 pandemic caused by the SARS-CoV-2 virus. The medical conditions associated with SARS-CoV-2 infections have resulted in a surge in the number of patients at clinics and hospitals, leading to a significantly increased strain on healthcare resources. As such, an important part of managing and handling patients with SARS-CoV-2 infections within the clinical workflow is severity assessment, which is often conducted with the use of chest x-ray (CXR) images. In this work, we introduce COVID-Net CXR-S, a convolutional neural network for predicting the airspace severity of a SARS-CoV-2 positive patient based on a CXR image of the patient's chest. More specifically, we leveraged transfer learning to transfer representational knowledge gained from over 16,000 CXR images from a multinational cohort of over 15,000 patient cases into a custom network architecture for severity assessment. Experimental results with a multi-national patient cohort curated by the Radiological Society of North America (RSNA) RICORD initiative showed that the proposed COVID-Net CXR-S has potential to be a powerful tool for computer-aided severity assessment of CXR images of COVID-19 positive patients. Furthermore, radiologist validation on select cases by two board-certified radiologists with over 10 and 19 years of experience, respectively, showed consistency between radiologist interpretation and critical factors leveraged by COVID-Net CXR-S for severity assessment. While not a production-ready solution, the ultimate goal for the open source release of COVID-Net CXR-S is to act as a catalyst for clinical scientists, machine learning researchers, as well as citizen scientists to develop innovative new clinical decision support solutions for helping clinicians around the world manage the continuing pandemic.

James Ren Hou Lee, Maya Pavlova, Mahmoud Famouri et al.

Skin cancer continues to be the most frequently diagnosed form of cancer in the U.S., with not only significant effects on health and well-being but also significant economic costs associated with treatment. A crucial step to the treatment and management of skin cancer is effective skin cancer detection due to strong prognosis when treated at an early stage, with one of the key screening approaches being dermoscopy examination. Motivated by the advances of deep learning and inspired by the open source initiatives in the research community, in this study we introduce CancerNet-SCa, a suite of deep neural network designs tailored for the detection of skin cancer from dermoscopy images that is open source and available to the general public as part of the Cancer-Net initiative. To the best of the authors' knowledge, CancerNet-SCa comprises of the first machine-designed deep neural network architecture designs tailored specifically for skin cancer detection, one of which possessing a self-attention architecture design with attention condensers. Furthermore, we investigate and audit the behaviour of CancerNet-SCa in a responsible and transparent manner via explainability-driven model auditing. While CancerNet-SCa is not a production-ready screening solution, the hope is that the release of CancerNet-SCa in open source, open access form will encourage researchers, clinicians, and citizen data scientists alike to leverage and build upon them.

Arman Zharmagambetov, Chuan Guo, Ivan Evtimov et al.

Autonomous AI agents that can follow instructions and perform complex multi-step tasks have tremendous potential to boost human productivity. However, to perform many of these tasks, the agents need access to personal information from their users, raising the question of whether they are capable of using it appropriately. In this work, we introduce a new benchmark AgentDAM that measures if AI web-navigation agents follow the privacy principle of ``data minimization''. For the purposes of our benchmark, data minimization means that the agent uses a piece of potentially sensitive information only if it is ``necessary'' to complete a particular task. Our benchmark simulates realistic web interaction scenarios end-to-end and is adaptable to all existing web navigation agents. We use AgentDAM to evaluate how well AI agents built on top of GPT-4, Llama-3 and Claude can limit processing of potentially private information, and show that they are prone to inadvertent use of unnecessary sensitive information. We also propose a prompting-based defense that reduces information leakage, and demonstrate that our end-to-end benchmarking provides a more realistic measure than probing LLMs about privacy. Our results highlight that further research is needed to develop AI agents that can prioritize data minimization at inference time.

Hossein Aboutalebi, Maya Pavlova, Hayden Gunraj et al.

Medical image analysis continues to hold interesting challenges given the subtle characteristics of certain diseases and the significant overlap in appearance between diseases. In this work, we explore the concept of self-attention for tackling such subtleties in and between diseases. To this end, we introduce MEDUSA, a multi-scale encoder-decoder self-attention mechanism tailored for medical image analysis. While self-attention deep convolutional neural network architectures in existing literature center around the notion of multiple isolated lightweight attention mechanisms with limited individual capacities being incorporated at different points in the network architecture, MEDUSA takes a significant departure from this notion by possessing a single, unified self-attention mechanism with significantly higher capacity with multiple attention heads feeding into different scales in the network architecture. To the best of the authors' knowledge, this is the first "single body, multi-scale heads" realization of self-attention and enables explicit global context amongst selective attention at different levels of representational abstractions while still enabling differing local attention context at individual levels of abstractions. With MEDUSA, we obtain state-of-the-art performance on multiple challenging medical image analysis benchmarks including COVIDx, RSNA RICORD, and RSNA Pneumonia Challenge when compared to previous work. Our MEDUSA model is publicly available.

Audrey G. Chung, Maya Pavlova, Hayden Gunraj et al.

As the COVID-19 pandemic continues to devastate globally, one promising field of research is machine learning-driven computer vision to streamline various parts of the COVID-19 clinical workflow. These machine learning methods are typically stand-alone models designed without consideration for the integration necessary for real-world application workflows. In this study, we take a machine learning and systems (MLSys) perspective to design a system for COVID-19 patient screening with the clinical workflow in mind. The COVID-Net system is comprised of the continuously evolving COVIDx dataset, COVID-Net deep neural network for COVID-19 patient detection, and COVID-Net S deep neural networks for disease severity scoring for COVID-19 positive patient cases. The deep neural networks within the COVID-Net system possess state-of-the-art performance, and are designed to be integrated within a user interface (UI) for clinical decision support with automatic report generation to assist clinicians in their treatment decisions.

Alexander Wong, Mahmoud Famouri, Maya Pavlova et al.

Advances in deep learning have led to state-of-the-art performance across a multitude of speech recognition tasks. Nevertheless, the widespread deployment of deep neural networks for on-device speech recognition remains a challenge, particularly in edge scenarios where the memory and computing resources are highly constrained (e.g., low-power embedded devices) or where the memory and computing budget dedicated to speech recognition is low (e.g., mobile devices performing numerous tasks besides speech recognition). In this study, we introduce the concept of attention condensers for building low-footprint, highly-efficient deep neural networks for on-device speech recognition on the edge. An attention condenser is a self-attention mechanism that learns and produces a condensed embedding characterizing joint local and cross-channel activation relationships, and performs selective attention accordingly. To illustrate its efficacy, we introduce TinySpeech, low-precision deep neural networks comprising largely of attention condensers tailored for on-device speech recognition using a machine-driven design exploration strategy, with one tailored specifically with microcontroller operation constraints. Experimental results on the Google Speech Commands benchmark dataset for limited-vocabulary speech recognition showed that TinySpeech networks achieved significantly lower architectural complexity (as much as $507\times$ fewer parameters), lower computational complexity (as much as $48\times$ fewer multiply-add operations), and lower storage requirements (as much as $2028\times$ lower weight memory requirements) when compared to previous work. These results not only demonstrate the efficacy of attention condensers for building highly efficient networks for on-device speech recognition, but also illuminate its potential for accelerating deep learning on the edge and empowering TinyML applications.