Research Papers

Multi-objective evolutionary algorithms (MOEAs) have emerged as powerful tools for solving complex optimization problems characterized by multiple, often conflicting, objectives. While advancements have been made in computational efficiency as well as diversity and convergence of solutions, a critical challenge persists: the internal evolutionary mechanisms are opaque to human users. Drawing upon the successes of explainable AI in explaining complex algorithms and models, we argue that the need to understand the underlying evolutionary operators and population dynamics within MOEAs aligns well with a visual analytics paradigm. This paper introduces ParetoTracker, a visual analytics framework designed to support the comprehension and inspection of population dynamics in the evolutionary processes of MOEAs. Informed by preliminary literature review and expert interviews, the framework establishes a multi-level analysis scheme, which caters to user engagement and exploration ranging from examining overall trends in performance metrics to conducting fine-grained inspections of evolutionary operations. In contrast to conventional practices that require manual plotting of solutions for each generation, ParetoTracker facilitates the examination of temporal trends and dynamics across consecutive generations in an integrated visual interface. The effectiveness of the framework is demonstrated through case studies and expert interviews focused on widely adopted benchmark optimization problems.

We introduce the Consistent Reasoning Paradox (CRP). Consistent reasoning, which lies at the core of human intelligence, is the ability to handle tasks that are equivalent, yet described by different sentences ('Tell me the time!' and 'What is the time?'). The CRP asserts that consistent reasoning implies fallibility -- in particular, human-like intelligence in AI necessarily comes with human-like fallibility. Specifically, it states that there are problems, e.g. in basic arithmetic, where any AI that always answers and strives to mimic human intelligence by reasoning consistently will hallucinate (produce wrong, yet plausible answers) infinitely often. The paradox is that there exists a non-consistently reasoning AI (which therefore cannot be on the level of human intelligence) that will be correct on the same set of problems. The CRP also shows that detecting these hallucinations, even in a probabilistic sense, is strictly harder than solving the original problems, and that there are problems that an AI may answer correctly, but it cannot provide a correct logical explanation for how it arrived at the answer. Therefore, the CRP implies that any trustworthy AI (i.e., an AI that never answers incorrectly) that also reasons consistently must be able to say 'I don't know'. Moreover, this can only be done by implicitly computing a new concept that we introduce, termed the 'I don't know' function -- something currently lacking in modern AI. In view of these insights, the CRP also provides a glimpse into the behaviour of Artificial General Intelligence (AGI). An AGI cannot be 'almost sure', nor can it always explain itself, and therefore to be trustworthy it must be able to say 'I don't know'.

Large Language Models (LLMs) are able to provide assistance on a wide range of information-seeking tasks. However, model outputs may be misleading, whether unintentionally or in cases of intentional deception. We investigate the ability of LLMs to be deceptive in the context of providing assistance on a reading comprehension task, using LLMs as proxies for human users. We compare outcomes of (1) when the model is prompted to provide truthful assistance, (2) when it is prompted to be subtly misleading, and (3) when it is prompted to argue for an incorrect answer. Our experiments show that GPT-4 can effectively mislead both GPT-3.5-Turbo and GPT-4, with deceptive assistants resulting in up to a 23% drop in accuracy on the task compared to when a truthful assistant is used. We also find that providing the user model with additional context from the passage partially mitigates the influence of the deceptive model. This work highlights the ability of LLMs to produce misleading information and the effects this may have in real-world situations.

SLM as Guardian: Pioneering AI Safety with Small Language Models

Ohjoon Kwon, Donghyeon Jeon, Nayoung Choi +6

Most prior safety research of large language models (LLMs) has focused on enhancing the alignment of LLMs to better suit the safety requirements of humans. However, internalizing such safeguard features into larger models brought challenges of higher training cost and unintended degradation of helpfulness. To overcome such challenges, a modular approach employing a smaller LLM to detect harmful user queries is regarded as a convenient solution in designing LLM-based system with safety requirements. In this paper, we leverage a smaller LLM for both harmful query detection and safeguard response generation. We introduce our safety requirements and the taxonomy of harmfulness categories, and then propose a multi-task learning mechanism fusing the two tasks into a single model. We demonstrate the effectiveness of our approach, providing on par or surpassing harmful query detection and safeguard response performance compared to the publicly available LLMs.

Recently, there has been growing attention on behalf of both academic and practice communities towards the ability of Artificial Intelligence (AI) systems to operate responsibly and ethically. As a result, a plethora of frameworks and guidelines have appeared to support practitioners in implementing Trustworthy AI applications (TAI). However, little research has been done to investigate whether such frameworks are being used and how. In this work, we study the vision AI practitioners have on TAI principles, how they address them, and what they would like to have - in terms of tools, knowledge, or guidelines - when they attempt to incorporate such principles into the systems they develop. Through a survey and semi-structured interviews, we systematically investigated practitioners' challenges and needs in developing TAI systems. Based on these practical findings, we highlight recommendations to help AI practitioners develop Trustworthy AI applications.

This paper explores the value of archival theory as a means of grappling with bias in algorithmic design. Rather than seek to mitigate biases perpetuated by datasets and algorithmic systems, archival theory offers a reframing of bias itself. Drawing on a range of archival theory from the fields of history, literary and cultural studies, Black studies, and feminist STS, we propose absence-as power, presence, and productive-as a concept that might more securely anchor investigations into the causes of algorithmic bias, and that can prompt more capacious, creative, and joyful future work. This essay, in turn, can intervene into the technical as well as the social, historical, and political structures that serve as sources of bias.

As the manufacturing industry advances with sensor integration and automation, the opaque nature of deep learning models in machine learning poses a significant challenge for fault detection and diagnosis. And despite the related predictive insights Artificial Intelligence (AI) can deliver, advanced machine learning engines often remain a black box. This paper reviews the eXplainable AI (XAI) tools and techniques in this context. We explore various XAI methodologies, focusing on their role in making AI decision-making transparent, particularly in critical scenarios where humans are involved. We also discuss current limitations and potential future research that aims to balance explainability with model performance while improving trustworthiness in the context of AI applications for critical industrial use cases.

Exploring Explainability in Video Action Recognition

Avinab Saha, Shashank Gupta, Sravan Kumar Ankireddy +2

Image Classification and Video Action Recognition are perhaps the two most foundational tasks in computer vision. Consequently, explaining the inner workings of trained deep neural networks is of prime importance. While numerous efforts focus on explaining the decisions of trained deep neural networks in image classification, exploration in the domain of its temporal version, video action recognition, has been scant. In this work, we take a deeper look at this problem. We begin by revisiting Grad-CAM, one of the popular feature attribution methods for Image Classification, and its extension to Video Action Recognition tasks and examine the method's limitations. To address these, we introduce Video-TCAV, by building on TCAV for Image Classification tasks, which aims to quantify the importance of specific concepts in the decision-making process of Video Action Recognition models. As the scalable generation of concepts is still an open problem, we propose a machine-assisted approach to generate spatial and spatiotemporal concepts relevant to Video Action Recognition for testing Video-TCAV. We then establish the importance of temporally-varying concepts by demonstrating the superiority of dynamic spatiotemporal concepts over trivial spatial concepts. In conclusion, we introduce a framework for investigating hypotheses in action recognition and quantitatively testing them, thus advancing research in the explainability of deep neural networks used in video action recognition.

Integrating artificial intelligence (AI) and federated learning (FL) in smart transportation has raised critical issues regarding their responsible use. Ensuring responsible AI is paramount for the stability and sustainability of intelligent transportation systems. Despite its importance, research on the responsible application of AI and FL in this domain remains nascent, with a paucity of in-depth investigations into their confluence. Our study analyzes the roles of FL in smart transportation, as well as the promoting effect of responsible AI on distributed smart transportation. Lastly, we discuss the challenges of developing and implementing responsible FL in smart transportation and propose potential solutions. By integrating responsible AI and federated learning, intelligent transportation systems are expected to achieve a higher degree of intelligence, personalization, safety, and transparency.

Wildfires present intricate challenges for prediction, necessitating the use of sophisticated machine learning techniques for effective modeling\cite{jain2020review}. In our research, we conducted a thorough assessment of various machine learning algorithms for both classification and regression tasks relevant to predicting wildfires. We found that for classifying different types or stages of wildfires, the XGBoost model outperformed others in terms of accuracy and robustness. Meanwhile, the Random Forest regression model showed superior results in predicting the extent of wildfire-affected areas, excelling in both prediction error and explained variance. Additionally, we developed a hybrid neural network model that integrates numerical data and image information for simultaneous classification and regression. To gain deeper insights into the decision-making processes of these models and identify key contributing features, we utilized eXplainable Artificial Intelligence (XAI) techniques, including TreeSHAP, LIME, Partial Dependence Plots (PDP), and Gradient-weighted Class Activation Mapping (Grad-CAM). These interpretability tools shed light on the significance and interplay of various features, highlighting the complex factors influencing wildfire predictions. Our study not only demonstrates the effectiveness of specific machine learning models in wildfire-related tasks but also underscores the critical role of model transparency and interpretability in environmental science applications.

Aurora-M: Open Source Continual Pre-training for Multilingual Language and Code

Taishi Nakamura, Mayank Mishra, Simone Tedeschi +42

Pretrained language models are an integral part of AI applications, but their high computational cost for training limits accessibility. Initiatives such as Bloom and StarCoder aim to democratize access to pretrained models for collaborative community development. Despite these efforts, such models encounter challenges such as limited multilingual capabilities, risks of catastrophic forgetting during continual pretraining, and the high costs of training models from scratch, alongside the need to align with AI safety standards and regulatory frameworks. This paper presents Aurora-M, a 15B parameter multilingual open-source model trained on English, Finnish, Hindi, Japanese, Vietnamese, and code. Continually pretrained from StarCoderPlus on 435B additional tokens, Aurora-M surpasses 2T tokens in total training token count. It is the first open-source multilingual model fine-tuned on human-reviewed safety instructions, thus aligning its development not only with conventional red-teaming considerations, but also with the specific concerns articulated in the Biden-Harris Executive Order on the Safe, Secure, and Trustworthy Development and Use of Artificial Intelligence. We evaluate Aurora-M across a wide range of tasks and languages, showcasing its robustness against catastrophic forgetting and its superior performance in multilingual settings, particularly in safety evaluations. We open-source Aurora-M and its variants to encourage responsible open-source development of large language models at https://huggingface.co/aurora-m.

Human-centered explainable AI (HCXAI) advocates for the integration of social aspects into AI explanations. Central to the HCXAI discourse is the Social Transparency (ST) framework, which aims to make the socio-organizational context of AI systems accessible to their users. In this work, we suggest extending the ST framework to address the risks of social misattributions in Large Language Models (LLMs), particularly in sensitive areas like mental health. In fact LLMs, which are remarkably capable of simulating roles and personas, may lead to mismatches between designers' intentions and users' perceptions of social attributes, risking to promote emotional manipulation and dangerous behaviors, cases of epistemic injustice, and unwarranted trust. To address these issues, we propose enhancing the ST framework with a fifth 'W-question' to clarify the specific social attributions assigned to LLMs by its designers and users. This addition aims to bridge the gap between LLM capabilities and user perceptions, promoting the ethically responsible development and use of LLM-based technology.