arXiv Computer Science @arxiv_cs@qoto.org

Responsible Artificial Intelligence (RAI) has emerged as a crucial framework for addressing ethical concerns in the development and deployment of Artificial Intelligence (AI) systems. A significant body of literature exists, primarily focusing on either RAI guidelines and principles or the technical aspects of RAI, largely within the realm of traditional AI. However, a notable gap persists in bridging theoretical frameworks with practical implementations in real-world settings, as well as transitioning from RAI to Responsible Generative AI (Gen AI). To bridge this gap, we present this article, which examines the challenges and opportunities in implementing ethical, transparent, and accountable AI systems in the post-ChatGPT era, an era significantly shaped by Gen AI. Our analysis includes governance and technical frameworks, the exploration of explainable AI as the backbone to achieve RAI, key performance indicators in RAI, alignment of Gen AI benchmarks with governance frameworks, reviews of AI-ready test beds, and RAI applications across multiple sectors. Additionally, we discuss challenges in RAI implementation and provide a philosophical perspective on the future of RAI. This comprehensive article aims to offer an overview of RAI, providing valuable insights for researchers, policymakers, users, and industry practitioners to develop and deploy AI systems that benefit individuals and society while minimizing potential risks and societal impacts. A curated list of resources and datasets covered in this survey is available on GitHub {https://github.com/anas-zafar/Responsible-AI}.

The development and deployment of artificial intelligence (AI) systems, with their profound societal impacts, raise critical challenges for governance. Historically, technological innovations have been governed by concentrated expertise with limited public input. However, AI's pervasive influence across domains such as healthcare, employment, and justice necessitates inclusive governance approaches. This article explores the tension between expert-led oversight and democratic participation, analyzing models of participatory and deliberative democracy. Using case studies from France and Brazil, we highlight how inclusive frameworks can bridge the gap between technical complexity and public accountability. Recommendations are provided for integrating these approaches into a balanced governance model tailored to the European Union, emphasizing transparency, diversity, and adaptive regulation to ensure that AI governance reflects societal values while maintaining technical rigor. This analysis underscores the importance of hybrid frameworks that unite expertise and public voice in shaping the future of AI policy.

This research introduces LegalScore, a specialized index for assessing how generative artificial intelligence models perform in a selected range of career exams that require a legal background in Brazil. The index evaluates fourteen different types of artificial intelligence models' performance, from proprietary to open-source models, in answering objective questions applied to these exams. The research uncovers the response of the models when applying English-trained large language models to Brazilian legal contexts, leading us to reflect on the importance and the need for Brazil-specific training data in generative artificial intelligence models. Performance analysis shows that while proprietary and most known models achieved better results overall, local and smaller models indicated promising performances due to their Brazilian context alignment in training. By establishing an evaluation framework with metrics including accuracy, confidence intervals, and normalized scoring, LegalScore enables systematic assessment of artificial intelligence performance in legal examinations in Brazil. While the study demonstrates artificial intelligence's potential value for exam preparation and question development, it concludes that significant improvements are needed before AI can match human performance in advanced legal assessments. The benchmark creates a foundation for continued research, highlighting the importance of local adaptation in artificial intelligence development.

The widespread adoption of large language models (LLMs) marks a transformative era in technology, especially within the educational sector. This paper explores the integration of LLMs within learning management systems (LMSs) to develop an adaptive learning management system (ALMS) personalized for individual learners across various educational stages. Traditional LMSs, while facilitating the distribution of educational materials, fall short in addressing the nuanced needs of diverse student populations, particularly in settings with limited instructor availability. Our proposed system leverages the flexibility of AI to provide a customizable learning environment that adjusts to each user's evolving needs. By integrating a suite of general-purpose and domain-specific LLMs, this system aims to minimize common issues such as factual inaccuracies and outdated information, characteristic of general LLMs like OpenAI's ChatGPT. This paper details the development of an ALMS that not only addresses privacy concerns and the limitations of existing educational tools but also enhances the learning experience by maintaining engagement through personalized educational content.

Recent AI agents, such as ChatGPT and LLaMA, primarily rely on instruction tuning and reinforcement learning to calibrate the output of large language models (LLMs) with human intentions, ensuring the outputs are harmless and helpful. Existing methods heavily depend on the manual annotation of high-quality positive samples, while contending with issues such as noisy labels and minimal distinctions between preferred and dispreferred response data. However, readily available toxic samples with clear safety distinctions are often filtered out, removing valuable negative references that could aid LLMs in safety alignment. In response, we propose PT-ALIGN, a novel safety self-alignment approach that minimizes human supervision by automatically refining positive and toxic samples and performing fine-grained dual instruction tuning. Positive samples are harmless responses, while toxic samples deliberately contain extremely harmful content, serving as a new supervisory signals. Specifically, we utilize LLM itself to iteratively generate and refine training instances by only exploring fewer than 50 human annotations. We then employ two losses, i.e., maximum likelihood estimation (MLE) and fine-grained unlikelihood training (UT), to jointly learn to enhance the LLM's safety. The MLE loss encourages an LLM to maximize the generation of harmless content based on positive samples. Conversely, the fine-grained UT loss guides the LLM to minimize the output of harmful words based on negative samples at the token-level, thereby guiding the model to decouple safety from effectiveness, directing it toward safer fine-tuning objectives, and increasing the likelihood of generating helpful and reliable content. Experiments on 9 popular open-source LLMs demonstrate the effectiveness of our PT-ALIGN for safety alignment, while maintaining comparable levels of helpfulness and usefulness.

Recently, artificial intelligence (AI)-enabled nonlinear vehicle platoon dynamics modeling plays a crucial role in predicting and optimizing the interactions between vehicles. Existing efforts lack the extraction and capture of vehicle behavior interaction features at the platoon scale. More importantly, maintaining high modeling accuracy without losing physical analyzability remains to be solved. To this end, this paper proposes a novel physics-encoded deep learning network, named PeMTFLN, to model the nonlinear vehicle platoon dynamics. Specifically, an analyzable parameters encoded computational graph (APeCG) is designed to guide the platoon to respond to the driving behavior of the lead vehicle while ensuring local stability. Besides, a multi-scale trajectory feature learning network (MTFLN) is constructed to capture platoon following patterns and infer the physical parameters required for APeCG from trajectory data. The human-driven vehicle trajectory datasets (HIGHSIM) were used to train the proposed PeMTFLN. The trajectories prediction experiments show that PeMTFLN exhibits superior compared to the baseline models in terms of predictive accuracy in speed and gap. The stability analysis result shows that the physical parameters in APeCG is able to reproduce the platoon stability in real-world condition. In simulation experiments, PeMTFLN performs low inference error in platoon trajectories generation. Moreover, PeMTFLN also accurately reproduces ground-truth safety statistics. The code of proposed PeMTFLN is open source.

Accurate and real-time prediction of surrounding vehicles' lane-changing intentions is a critical challenge in deploying safe and efficient autonomous driving systems in open-world scenarios. Existing high-performing methods remain hard to deploy due to their high computational cost, long training times, and excessive memory requirements. Here, we propose an efficient lane-changing intention prediction approach based on brain-inspired Spiking Neural Networks (SNN). By leveraging the event-driven nature of SNN, the proposed approach enables us to encode the vehicle's states in a more efficient manner. Comparison experiments conducted on HighD and NGSIM datasets demonstrate that our method significantly improves training efficiency and reduces deployment costs while maintaining comparable prediction accuracy. Particularly, compared to the baseline, our approach reduces training time by 75% and memory usage by 99.9%. These results validate the efficiency and reliability of our method in lane-changing predictions, highlighting its potential for safe and efficient autonomous driving systems while offering significant advantages in deployment, including reduced training time, lower memory usage, and faster inference.

Semantic role labeling (SRL) is a central natural language processing (NLP) task aiming to understand the semantic roles within texts, facilitating a wide range of downstream applications. While SRL has garnered extensive and enduring research, there is currently a lack of a comprehensive survey that thoroughly organizes and synthesizes the field. This paper aims to review the entire research trajectory of the SRL community over the past two decades. We begin by providing a complete definition of SRL. To offer a comprehensive taxonomy, we categorize SRL methodologies into four key perspectives: model architectures, syntax feature modeling, application scenarios, and multi-modal extensions. Further, we discuss SRL benchmarks, evaluation metrics, and paradigm modeling approaches, while also exploring practical applications across various domains. Finally, we analyze future research directions in SRL, addressing the evolving role of SRL in the age of large language models (LLMs) and its potential impact on the broader NLP landscape. We maintain a public repository and consistently update related resources at: https://github.com/DreamH1gh/Awesome-SRL

Automated Program Repair (APR) for introductory programming assignments (IPAs) is motivated by the large number of student enrollments in programming courses each year. Since providing feedback on IPAs requires substantial time and effort from faculty, personalized feedback often involves suggesting fixes to students' programs. Formal Methods (FM)-based semantic repair approaches, check a program's execution against a test suite or reference solution, are effective but limited. These tools excel at identifying buggy parts but can only fix programs if the correct implementation and the faulty one share the same control flow graph. Conversely, Large Language Models (LLMs) are used for APR but often make extensive instead of minimal rewrites. This leads to more invasive fixes, making it harder for students to learn from their mistakes. In summary, LLMs excel at completing strings, while FM-based fault localization excel at identifying buggy parts of a program. In this paper, we propose a novel approach that combines the strengths of both FM-based fault localization and LLMs, via zero-shot learning, to enhance APR for IPAs. Our method uses MaxSAT-based fault localization to identify buggy parts of a program, then presents the LLM with a program sketch devoid of these buggy statements. This hybrid approach follows a CEGIS loop to iteratively refine the program. We ask the LLM to synthesize the missing parts, which are then checked against a test suite. If the suggested program is incorrect, a counterexample from the test suite is fed back to the LLM. Our experiments show that our counterexample guided approach, using MaxSAT-based bug-free program sketches, significantly improves the repair capabilities of all six evaluated LLMs. This method allows LLMs to repair more programs with smaller fixes, outperforming other configurations and state-of-the-art symbolic program repair tools.

Efficient and socially equitable restoration of transportation networks post disasters is crucial for community resilience and access to essential services. The ability to rapidly recover critical infrastructure can significantly mitigate the impacts of disasters, particularly in underserved communities where prolonged isolation exacerbates vulnerabilities. Traditional restoration methods prioritize functionality over computational efficiency and equity, leaving low-income communities at a disadvantage during recovery. To address this gap, this research introduces a novel framework that combines quantum computing technology with an equity-focused approach to network restoration. Optimization of road link recovery within budget constraints is achieved by leveraging D Wave's hybrid quantum solver, which targets the connectivity needs of low, average, and high income communities. This framework combines computational speed with equity, ensuring priority support for underserved populations. Findings demonstrate that this hybrid quantum solver achieves near instantaneous computation times of approximately 8.7 seconds across various budget scenarios, significantly outperforming the widely used genetic algorithm. It offers targeted restoration by first aiding low-income communities and expanding aid as budgets increase, aligning with equity goals. This work showcases quantum computing's potential in disaster recovery planning, providing a rapid and equitable solution that elevates urban resilience and social sustainability by aiding vulnerable populations in disasters.

The recent explosion of "foundation" generative AI models has been built upon the extensive extraction of value from online sources, often without corresponding reciprocation. This pattern mirrors and intensifies the extractive practices of surveillance capitalism, while the potential for enormous profit has challenged technology organizations' commitments to responsible AI practices, raising significant ethical and societal concerns. However, a promising alternative is emerging: the development of models that rely on content willingly and collaboratively provided by users. This article explores this "egalitarian" approach to generative AI, taking inspiration from the successful model of Wikipedia. We explore the potential implications of this approach for the design, development, and constraints of future foundation models. We argue that such an approach is not only ethically sound but may also lead to models that are more responsive to user needs, more diverse in their training data, and ultimately more aligned with societal values. Furthermore, we explore potential challenges and limitations of this approach, including issues of scalability, quality control, and potential biases inherent in volunteer-contributed content.

Numerical modelling of coupled multiphysics phenomena is becoming an increasingly important subject in applied mathematics. The main challenge in teaching this subject is the complexity of both the mathematical models and their numerical implementation. In this note, a simple demonstrator is proposed that enables demonstration of and some hands-on experience with three types of coupling commonly used in computational science: one-way coupling, explicit sequential coupling, and full coupling. It makes use of a familiar nonlinear heat equation in 1D, with the solution being a propagating heat wave.

There is a range of ways to organize hiring and admissions in higher education, as in many domains, ranging from very centralized processes where a single person makes final decisions to very decentralized processes where many people make decisions about who to admit or hire. Decentralized processes can enable individual and collective empowerment, but this may come at the cost of efficiency. With the advent of automated decision making, this question of centralization has a big impact on hiring and admissions, given that automated systems often are easier to implement, or even require, more centralized decision making. In this paper, we develop a strategic model to explore the impact of the degree of centralization on both the candidates and the hirers, with a focus on university admissions. The model reflects a trade-off between a centralized committee where preferences may not capture individual hirers' preferences, and a decentralized process where individual hirers face extra costs to interview candidates themselves. We characterize when individual hirers prefer the decentralized process over the centralized process as a function of the degree to which the centralized process and hirers' preferences are aligned. We also show that decentralization can have devastating consequences for fairness, leading to major disparities in the likelihood of getting hired across candidates. Our results demonstrate the trade-offs that occur under the question of centralization vs decentralization, and point to how an answer to this question can impose significant harm to people in these systems.

The integration of generative AI (GenAI) and large language models (LLMs) in healthcare presents both unprecedented opportunities and challenges, necessitating innovative regulatory approaches. GenAI and LLMs offer broad applications, from automating clinical workflows to personalizing diagnostics. However, the non-deterministic outputs, broad functionalities and complex integration of GenAI and LLMs challenge existing medical device regulatory frameworks, including the total product life cycle (TPLC) approach. Here we discuss the constraints of the TPLC approach to GenAI and LLM-based medical device regulation, and advocate for global collaboration in regulatory science research. This serves as the foundation for developing innovative approaches including adaptive policies and regulatory sandboxes, to test and refine governance in real-world settings. International harmonization, as seen with the International Medical Device Regulators Forum, is essential to manage implications of LLM on global health, including risks of widening health inequities driven by inherent model biases. By engaging multidisciplinary expertise, prioritizing iterative, data-driven approaches, and focusing on the needs of diverse populations, global regulatory science research enables the responsible and equitable advancement of LLM innovations in healthcare.

This paper introduces an efficient stabilizer-free weak Galerkin (WG) finite element method for solving the three-dimensional quad-curl problem. Leveraging bubble functions as a key analytical tool, the method extends the applicability of stabilizer-free WG approaches to non-convex elements in finite element partitions-a notable advancement over existing methods, which are restricted to convex elements. The proposed method maintains a simple, symmetric, and positive definite formulation. It achieves optimal error estimates for the exact solution in a discrete norm, as well as an optimal-order $L^2$ error estimate for $k>2$ and a sub-optimal order for the lowest order case $k=2$. Numerical experiments are presented to validate the method's efficiency and accuracy.

This paper develops an efficient computational technique to assess the landslide responses to tectonic deformation and to predict the implications of large bedrocks landslides on the short and long-term development of the disasters. The considered equations represent a three-dimensional system of geological structure deformation subject to suitable initial and boundary conditions. The space derivatives are approximated using the finite element procedure while the approximation in time derivative is obtained using the Lax-Wendroff and interpolation techniques. The new approach is so called a combined Lax-Wendroff/interpolation method with finite element method. The modified Lax-Wendroff/interpolation scheme is employed to efficiently treat the time derivative term and to provide a suitable time step restriction for stability. Under this time step requirement, both stability and error estimates of the new algorithm are deeply analyzed using a constructed strong norm. The theory suggests that the developed computational technique is second-order accurate in time and spatial convergent with order O(h^{p}), where $h$ denotes the space size and p is a positive integer. A wide set of numerical examples are carried out to confirm the theoretical results and to demonstrate the utility and validity of the proposed numerical scheme. An application to landslides observed in west and center regions in Cameroon from October 2019 to November 2024, are discussed.

This paper deals with the scheme proposed by the authors in Zorío, Baeza and Mulet (J Sci Comput 71(1):246-273, 2017). This scheme is an alternative to the techniques proposed in Qiu and Shu (SIAM J Sci Comput 24(6):2185-2198, 2003) to obtain high-order accurate schemes using Weighted Essentially Non Oscillatory finite differences and approximating the flux derivatives required by the Cauchy-Kovalevskaya procedure by simple centered finite differences. We analyse how errors in first-order terms near discontinuities propagate through both versions of the Cauchy-Kovalevskaya procedure. We propose a fluctuation control, for which the approximation of the first-order derivative to be used in the Cauchy-Kovalevskaya procedure is obtained from a Weighted Essentially Non Oscillatory (WENO) interpolation of flux derivatives, instead of the usual finite difference of WENO flux reconstructions. The numerical results that we obtain confirm the benefits of this fluctuation control.

Interpreting the internal mechanisms of large language models (LLMs) is crucial for improving their trustworthiness and utility. Prior work has primarily focused on mapping individual neurons to discrete semantic concepts. However, such mappings struggle to handle the inherent polysemanticity in LLMs, where individual neurons encode multiple, distinct concepts. Through a comprehensive analysis of both encoder and decoder-based LLMs across diverse datasets, we observe that even highly salient neurons, identified via various attribution techniques for specific semantic concepts, consistently exhibit polysemantic behavior. Importantly, activation magnitudes for fine-grained concepts follow distinct, often Gaussian-like distributions with minimal overlap. This observation motivates a shift from neuron attribution to range-based interpretation. We hypothesize that interpreting and manipulating neuron activation ranges would enable more precise interpretability and targeted interventions in LLMs. To validate our hypothesis, we introduce NeuronLens, a novel range-based interpretation and manipulation framework that provides a finer view of neuron activation distributions to localize concept attribution within a neuron. Extensive empirical evaluations demonstrate that NeuronLens significantly reduces unintended interference, while maintaining precise manipulation of targeted concepts, outperforming neuron attribution.