Search

Large Language Models (LLMs) are the cornerstones of modern artificial intelligence systems. This paper introduces Juhaina, a Arabic-English bilingual LLM specifically designed to align with the values and preferences of Arabic speakers. Juhaina inherently supports advanced functionalities such as instruction following, open-ended question answering, information provisioning, and text processing. Our model contains 9.24 billion parameters and is trained on a context window of up to 8,192 tokens. This paper details the creation process of Juhaina and provides an extensive empirical evaluation. Furthermore, we identify the limitations of widely-adopted Open Arabic LLM Leaderboard (OALL) and propose a new evaluation benchmark, CamelEval. Our findings demonstrate that Juhaina surpasses existing LLMs of comparable sizes, such as the Llama and Gemma families, in generating helpful responses in Arabic, providing factually accurate information about the region, and understanding nuanced cultural aspects. We aspire for Juhaina to democratize cutting-edge AI technologies, serving over 400 million Arabic speakers by offering LLMs that not only communicate in their language but also comprehend their culture. We publicly release all models on Huggingface \url{https://huggingface.co/elmrc}.

Deep learning has achieved outstanding accuracy in medical image segmentation, particularly for objects like organs or tumors with smooth boundaries or large sizes. Whereas, it encounters significant difficulties with objects that have zigzag boundaries or are small in size, leading to a notable decrease in segmentation effectiveness. In this context, using a loss function that incorporates smoothness and volume information into a model's predictions offers a promising solution to these shortcomings. In this work, we introduce an Adaptive Focal Loss (A-FL) function designed to mitigate class imbalance by down-weighting the loss for easy examples that results in up-weighting the loss for hard examples and giving greater emphasis to challenging examples, such as small and irregularly shaped objects. The proposed A-FL involves dynamically adjusting a focusing parameter based on an object's surface smoothness, size information, and adjusting the class balancing parameter based on the ratio of targeted area to total area in an image. We evaluated the performance of the A-FL using ResNet50-encoded U-Net architecture on the Picai 2022 and BraTS 2018 datasets. On the Picai 2022 dataset, the A-FL achieved an Intersection over Union (IoU) of 0.696 and a Dice Similarity Coefficient (DSC) of 0.769, outperforming the regular Focal Loss (FL) by 5.5% and 5.4% respectively. It also surpassed the best baseline Dice-Focal by 2.0% and 1.2%. On the BraTS 2018 dataset, A-FL achieved an IoU of 0.883 and a DSC of 0.931. The comparative studies show that the proposed A-FL function surpasses conventional methods, including Dice Loss, Focal Loss, and their hybrid variants, in IoU, DSC, Sensitivity, and Specificity metrics. This work highlights A-FL's potential to improve deep learning models for segmenting clinically significant regions in medical images, leading to more precise and reliable diagnostic tools., Comment: 15 pages, 4 figures

Deterministic policies are often preferred over stochastic ones when implemented on physical systems. They can prevent erratic and harmful behaviors while being easier to implement and interpret. However, in practice, exploration is largely performed by stochastic policies. First-order Bayesian Optimization (BO) methods offer a principled way of performing exploration using deterministic policies. This is done through a learned probabilistic model of the objective function and its gradient. Nonetheless, such approaches treat policy search as a black-box problem, and thus, neglect the reinforcement learning nature of the problem. In this work, we leverage the performance difference lemma to introduce a novel mean function for the probabilistic model. This results in augmenting BO methods with the action-value function. Hence, we call our method Augmented Bayesian Search~(ABS). Interestingly, this new mean function enhances the posterior gradient with the deterministic policy gradient, effectively bridging the gap between BO and policy gradient methods. The resulting algorithm combines the convenience of the direct policy search with the scalability of reinforcement learning. We validate ABS on high-dimensional locomotion problems and demonstrate competitive performance compared to existing direct policy search schemes., Comment: Accepted to the International Conference on Learning Representations (ICLR) 2024

Recently, huge uniaxial negative thermal expansion (NTE) along a $c$-axis has been observed in transition-metal ($Tr$) zirconides $Tr$Zr$_{2}$ with a tetragonal CuAl$_{2}$-type structure. In a recent study on FeZr$_{2}$ [M. Xu et al., Nat. Commun. 14, 4439 (2023)], the importance of optical phonons to the emergence of the $c$-axis NTE in FeZr$_{2}$ has been proposed. In this study, the physical properties of $Tr$Zr$_{2}$ ($Tr$ = Fe and Co$_{1-x}$Ni$_{x}$) have been studied by specific heat, sound velocity measurements, and theoretical phonon calculations to discuss the importance of optical phonons to the emergence of the $c$-axis NTE in CoZr$_{2}$ and FeZr$_{2}$. From analyses of lattice specific heat, we found that Ni substitution results in a systematic decrease in oscillator strength for the Einstein modes with 8.74 meV (CoZr$_{2}$). From phonon calculations, the low-energy optical phonon branches at the $\Gamma$ point were observed for CoZr$_{2}$ and FeZr$_{2}$ with $c$-axis NTE, but not in NiZr$_{2}$ with positive thermal expansion. The enhancement of phonon density of states near the above-mentioned optical phonon energy in CoZr$_{2}$ and FeZr$_{2}$ is consistent with the specific heat analyses. We propose the importance of the low-energy optical phonons to the emergence of the $c$-axis NTE in TrZr$_{2}$., Comment: 14 pages, 9 figures, and 3 tables

Requirement traceability is the process of identifying the inter-dependencies between requirements. It poses a significant challenge when conducted manually, especially when dealing with requirements at various levels of abstraction. In this work, we propose a novel approach to automate the task of linking high-level business requirements with more technical system requirements. The proposed approach begins by representing each requirement using a Bag of-Words (BOW) model combined with the Term Frequency-Inverse Document Frequency (TF-IDF) scoring function. Then, we suggested an enhanced cosine similarity that uses recent advances in word embedding representation to correct traditional cosine similarity function limitations. To evaluate the effectiveness of our approach, we conducted experiments on three well-known datasets: COEST, WARC(NFR), and WARC(FRS). The results demonstrate that our approach significantly improves efficiency compared to existing methods. We achieved better results with an increase of approximately 18.4% in one of the datasets, as measured by the F2 score.

Deep reinforcement learning agents are prone to goal misalignments. The black-box nature of their policies hinders the detection and correction of such misalignments, and the trust necessary for real-world deployment. So far, solutions learning interpretable policies are inefficient or require many human priors. We propose INTERPRETER, a fast distillation method producing INTerpretable Editable tRee Programs for ReinforcEmenT lEaRning. We empirically demonstrate that INTERPRETER compact tree programs match oracles across a diverse set of sequential decision tasks and evaluate the impact of our design choices on interpretability and performances. We show that our policies can be interpreted and edited to correct misalignments on Atari games and to explain real farming strategies.

Magnetic flux trapping in field-cooled (FC) Sn-Pb solders has been recently studied because of the observation of nonvolatile magneto-thermal switching [H. Arima et al., Commun. Mater. 5, 34 (2024)] and anomalous magnetic field-temperature (H-T) phase diagrams [T. Murakami et al., AIP Adv. 13, 125008 (2023)]. In this paper, we investigate the origin of the anomalously low specific heat (C) in Sn10-Pb90 and Sn45-Pb55 solders after FC at H = 1500 Oe. We show that the FC solders exhibit self-heating possibly caused by the flux flow during the reduction of trapped fluxes when heating the sample during the C measurements. The T dependence of T rise clearly exhibits unexpectedly large values when the low-C states are observed. In addition, the cause of the transition-like behavior in C-T of FC solders are explained by local heating during H control and flux-jump phenomena., Comment: 19 pages, 7 figures, Supplemental data. arXiv admin note: text overlap with arXiv:2403.01719

Formation of vortices is a typical phenomenon of type-II superconductors under magnetic fields (H). In contrast, type-I superconductors do not host vortices because of their Meissner state. As rare cases, vortices have been observed in intermediate states of type-I superconductors; in addition, the recent observation of type-II superconductivity with vortices in a (originally type-I) Pb crystal film at extremely low temperature (T) under H has opened new pathway to study vortex physics. However, thermodynamic characteristics of such type-II-like superconducting states with vortices in originally type-I element superconductors have not been detected because of the lack of bulk example. In this study, we investigated superconducting states of phase-separated Sn-Pb solders using specific heat and magnetization to reveal magnetic-flux-trapping mechanisms. Here, we show that Sn islands in the Sn-Pb solders exhibit type-II-like superconducting states with vortices when the solders host extremely high magnetic fluxes after field cooling. Furthermore, with increasing T, the amount of trapped flux decreases, and the driving force of the magnetic flux changes from type-II superconducting states of Sn to supercurrent of Pb regions surrounding the Sn islands. The field-cooled Sn-Pb solders are rich in physics of bulk vortex formation and anomalously enhanced supercurrent in element (originally type-I) superconductors., Comment: We found a problem on the analysis of specific heat data because of possible self-heating of the measured solders when flux reduction. We will resubmit the data after careful analyses

The classical Hertzian contact model establishes a monotonic correlation between contact force and area. Here, we showed that the interplay between local friction and structural instability can deliberately lead to unconventional contact behavior when a soft elastic shell comes into contact with a flat surface. The deviation from Hertzian contact first arises from bending within the contact area, followed by the second transition induced by buckling, resulting in a notable decrease in the contact area despite increased contact force. Friction delays both transitions and introduces hysteresis during unloading. However, a high amount of friction suppresses both buckling and dissipation. Different contact regimes are discussed in terms of rolling and sliding mechanisms, providing insights for tailoring contact behaviors in soft shells.

Value-alignment in normative multi-agent systems is used to promote a certain value and to ensure the consistent behaviour of agents in autonomous intelligent systems with human values. However, the current literature is limited to the incorporation of effective norms for single-value alignment with no consideration of agents' heterogeneity and the requirement of simultaneous promotion and alignment of multiple values. This research proposes a multi-value promotion model that uses multi-objective evolutionary algorithms and decentralised reasoning to produce the optimum parametric set of norms that is aligned with multiple simultaneous values of heterogeneous agents and the system. To understand various aspects of this complex problem, several evolutionary algorithms were used to find a set of optimised norm parameters considering two toy tax scenarios with two and five values are considered. The results are analysed from different perspectives to show the impact of a selected evolutionary algorithm on the solution, and the importance of understanding the relation between values when prioritising them., Comment: Accepted in MODeM 2023 Workshop at ECAI 2023. arXiv admin note: substantial text overlap with arXiv:2305.07366

The UK continues to face a huge adult-literacy challenge, with a sizeable percentage of the UK adult population having very low literacy skills. In addition to highlighting the literacy challenge in the UK, this report shows these individuals experience negative impacts on personal relationships, wellbeing, health, and education, as well as a greater risk of unemployment or being in low-paid work. In addition to many people struggling with low literacy skills, the report also showed that there are also many negative perceptions around, and attitudes towards, literacy. If more action is not taken to tackle low literacy in adults, there is the risk of perpetuating and reinforcing an intergenerational cycle of disadvantage.

Interpretability of AI models allows for user safety checks to build trust in such AIs. In particular, Decision Trees (DTs) provide a global look at the learned model and transparently reveal which features of the input are critical for making a decision. However, interpretability is hindered if the DT is too large. To learn compact trees, a recent Reinforcement Learning (RL) framework has been proposed to explore the space of DTs using deep RL. This framework augments a decision problem (e.g. a supervised classification task) with additional actions that gather information about the features of an otherwise hidden input. By appropriately penalizing these actions, the agent learns to optimally trade-off size and performance of DTs. In practice, a reactive policy for a partially observable Markov decision process (MDP) needs to be learned, which is still an open problem. We show in this paper that deep RL can fail even on simple toy tasks of this class. However, when the underlying decision problem is a supervised classification task, we show that finding the optimal tree can be cast as a fully observable Markov decision problem and be solved efficiently, giving rise to a new family of algorithms for learning DTs that go beyond the classical greedy maximization ones., Comment: To be included in an other submission. arXiv admin note: text overlap with arXiv:2304.05839

Finding an optimal decision tree for a supervised learning task is a challenging combinatorial problem to solve at scale. It was recently proposed to frame the problem as a Markov Decision Problem (MDP) and use deep reinforcement learning to tackle scaling. Unfortunately, these methods are not competitive with the current branch-and-bound state-of-the-art. We propose instead to scale the resolution of such MDPs using an information-theoretic tests generating function that heuristically, and dynamically for every state, limits the set of admissible test actions to a few good candidates. As a solver, we show empirically that our algorithm is at the very least competitive with branch-and-bound alternatives. As a machine learning tool, a key advantage of our approach is to solve for multiple complexity-performance trade-offs at virtually no additional cost. With such a set of solutions, a user can then select the tree that generalizes best and which has the interpretability level that best suits their needs, which no current branch-and-bound method allows.

Deep Attractor Network (DANet) is the state-of-the-art technique in speech separation field, which uses Bidirectional Long Short-Term Memory (BLSTM), but the complexity of the DANet model is very high. In this paper, a simplified and powerful DANet model is proposed using Bidirectional Gated neural network (BGRU) instead of BLSTM. The Gaussian Mixture Model (GMM) other than the k-means was applied in DANet as a clustering algorithm to reduce the complexity and increase the learning speed and accuracy. The metrics used in this paper are Signal to Distortion Ratio (SDR), Signal to Interference Ratio (SIR), Signal to Artifact Ratio (SAR), and Perceptual Evaluation Speech Quality (PESQ) score. Two speaker mixture datasets from TIMIT corpus were prepared to evaluate the proposed model, and the system achieved 12.3 dB and 2.94 for SDR and PESQ scores respectively, which were better than the original DANet model. Other improvements were 20.7% and 17.9% in the number of parameters and time training, respectively. The model was applied on mixed Arabic speech signals and the results were better than that in English.

Applying a magnetic field to a solid changes its thermal-transport properties. Although such magneto-thermal-transport phenomena are usually small effects, giant magneto-thermal resistance has recently been observed in spintronic materials1,2 and superconductors3,4, opening up new possibilities in thermal management technologies. However, the thermal conductivity conventionally changes only when a magnetic field is applied due to the absence of nonvolatility, which limits potential applications of thermal switching devices5,6. Here, we report the observation of nonvolatile thermal switching that changes the thermal conductivity when a magnetic field is applied and retains the value even when the field is turned off. This unconventional magneto-thermal switching, surprisingly, arises in commercial Sn-Pb solders and is realized by phase-separated superconducting states and resultant nonuniform magnetic flux distributions. This result confirms the versatility of the observed phenomenon and aids the development of active solid-state thermal management devices., Comment: 33 pages, 5 figures & 9 extended data figures