Face recognition is known to be vulnerable to adversarial face images. Existing works craft face adversarial images by indiscriminately changing a single attribute without being aware of the intrinsic attributes of the images. To this end, we propose a new Semantic Adversarial Attack called SAA-StarGAN that tampers with the significant facial attributes for each image. We predict the most significant attributes by applying the cosine similarity or probability score. The probability score method is based on training a Face Verification model for an attribute prediction task to obtain a class probability score for each attribute. The prediction process will help craft adversarial face images more easily and efficiently, as well as improve the adversarial transferability. Then, we change the most significant facial attributes, with either one or more of the facial attributes for impersonation and dodging attacks in white-box and black-box settings. Experimental results show that our method could generate diverse and realistic adversarial face images meanwhile avoid affecting human perception of the face recognition. SAA-StarGAN achieves an 80.5% attack success rate against black-box models, outperforming existing methods by 35.5% under the impersonation attack. Concerning the black-box setting, SAA-StarGAN achieves high attack success rates on various models. The experiments confirm that predicting the most important attributes significantly affects the success of adversarial attacks in both white-box and black-box settings and could enhance the transferability of the crafted adversarial examples.

The Coronavirus Disease 2019 or the COVID-19 pandemic has swept almost all parts of the world since the first case was found in Wuhan, China, in December 2019. With the increasing number of COVID-19 cases in the world, SARS-CoV-2 has mutated into various variants. Given the increasingly dangerous conditions of the pandemic, it is crucial to know when the pandemic will stop by predicting confirmed cases of COVID-19. Therefore, many studies have raised COVID-19 as a case study to overcome the ongoing pandemic using the Deep Learning method, namely LSTM, with reasonably accurate results and small error values. LSTM training is used to predict confirmed cases of COVID-19 based on variants that have been identified using ECDC's COVID-19 dataset containing confirmed cases of COVID-19 that have been identified from 30 countries in Europe. Tests were conducted using the LSTM and BiLSTM models with the addition of RNN as comparisons on hidden size and layer size. The obtained result showed that in testing hidden sizes 25, 50, 75 to 100, the RNN model provided better results, with the minimum MSE value of 0.01 and the RMSE value of 0.012 for B.1.427/B.1.429 variant with hidden size 100. In further testing of layer sizes 2, 3, 4, and 5, the result shows that the BiLSTM model provided better results, with minimum MSE value of 0.01 and the RMSE of 0.01 for the B.1.427/B.1.429 variant with hidden size 100 and layer size 2.

Review summarization is a non-trivial task that aims to summarize the main idea of the product review in the E-commerce website. Different from the document summary which only needs to focus on the main facts described in the document, review summarization should not only summarize the main aspects mentioned in the review but also reflect the personal style of the review author. Although existing review summarization methods have incorporated the historical reviews of both customer and product, they usually simply concatenate and indiscriminately model this two heterogeneous information into a long sequence. Moreover, the rating information can also provide a high-level abstraction of customer preference, it has not been used by the majority of methods. In this paper, we propose the Heterogeneous Historical Review aware Review Summarization Model (HHRRS) which separately models the two types of historical reviews with the rating information by a graph reasoning module with a contrastive loss. We employ a multi-task framework that conducts the review sentiment classification and summarization jointly. Extensive experiments on four benchmark datasets demonstrate the superiority of HHRRS on both tasks.

Deep learning methods dominate short-term high-resolution precipitation nowcasting in terms of prediction error. However, their operational usability is limited by difficulties explaining dynamics behind the predictions, which are smoothed out and missing the high-frequency features due to optimizing for mean error loss functions. We experiment with hand-engineering of the advection-diffusion differential equation into a PhyCell to introduce more accurate physical prior to a PhyDNet model that disentangles physical and residual dynamics. Results indicate that while PhyCell can learn the intended dynamics, training of PhyDNet remains driven by loss optimization, resulting in a model with the same prediction capabilities.

Variational autoencoders (VAEs) are a popular framework for modeling complex data distributions; they can be efficiently trained via variational inference by maximizing the evidence lower bound (ELBO), at the expense of a gap to the exact (log-)marginal likelihood. While VAEs are commonly used for disentangled representation learning, it is unclear why ELBO maximization would yield such representations, since unregularized maximum likelihood estimation generally cannot invert the data-generating process without additional assumptions. Yet, VAEs often succeed at this task. We seek to elucidate this apparent paradox by studying nonlinear VAEs in the limit of near-deterministic decoders. We first prove that, in this regime, the optimal encoder approximately inverts the decoder--a commonly used but unproven conjecture--which we refer to as self-consistency. Leveraging self-consistency, we show that the ELBO converges to a regularized log-likelihood. This allows VAEs to perform what has recently been termed independent mechanism analysis (IMA): it adds an inductive bias towards decoders with column-orthogonal Jacobians, which helps recovering the true latent factors. The gap between ELBO and log-likelihood is therefore welcome, since it bears unanticipated benefits for nonlinear representation learning. In experiments on synthetic and image data, we show that VAEs uncover the true latent factors when the data generating process satisfies the IMA assumption.

University of Stirling, United Kingdom vanicka.arora@stir.ac.uk Abstract As AI technologies are rolled out into healthcare, academia, human resources, law, and a multitude of other domains, they become de-facto arbiters of truth. But truth is highly contested, with many different definitions and approaches. It then investigates the production of truth in InstructGPT, a large language model, highlighting how data harvesting, model architectures, and social feedback mechanisms weave together disparate understandings of veracity. It conceptualizes this performance as an operationalization of truth, where distinct, often conflicting claims are smoothly synthesized and confidently presented into truth-statements. We argue that these same logics and inconsistencies play out in Instruct's successor, ChatGPT, reiterating truth as a non-trivial problem. We suggest that enriching sociality and thickening "reality" are two promising vectors for enhancing the truth-evaluating capacities of future language models. We conclude, however, by stepping back to consider AI truth-telling as a social practice: what kind of "truth" do we as listeners desire? OpenAI's latest language model appeared to We stress then that truth in AI is not just technical but be powerful and almost magical, generating news articles, also social, cultural, and political, drawing on particular writing poetry, and explaining arcane concepts norms and values. But a week later, the coding the technical matters: translating truth theories into site StackOverflow banned all answers produced actionable architectures and processes updates them by the model. "The primary problem," explained in significant ways. These disparate sociotechnical the staff, "is that while the answers which ChatGPT forces coalesce into a final AI model which purports produces have a high rate of being incorrect, they typically to tell the truth--and in doing so, our understanding look like they might be good and the answers of "truth" is remade. "The ideal of truth is a fallacy are very easy to produce" (Vincent 2022). For a site for semantic interpretation and needs to be changed," aiming to provide correct answers to coding problems, suggested two AI researchers (Welty and Aroyo 2015).

In this article, we use probing to investigate phenomena that occur during fine-tuning and knowledge distillation of a BERT-based natural language understanding (NLU) model. Our ultimate purpose was to use probing to better understand practical production problems and consequently to build better NLU models. We designed experiments to see how fine-tuning changes the linguistic capabilities of BERT, what the optimal size of the fine-tuning dataset is, and what amount of information is contained in a distilled NLU based on a tiny Transformer. The results of the experiments show that the probing paradigm in its current form is not well suited to answer such questions. Structural, Edge and Conditional probes do not take into account how easy it is to decode probed information. Consequently, we conclude that quantification of information decodability is critical for many practical applications of the probing paradigm.

Novel class discovery (NCD) aims to infer novel categories in an unlabeled dataset leveraging prior knowledge of a labeled set comprising disjoint but related classes. Existing research focuses primarily on utilizing the labeled set at the methodological level, with less emphasis on the analysis of the labeled set itself. Thus, in this paper, we rethink novel class discovery from the labeled set and focus on two core questions: (i) Given a specific unlabeled set, what kind of labeled set can best support novel class discovery? (ii) A fundamental premise of NCD is that the labeled set must be related to the unlabeled set, but how can we measure this relation? For (i), we propose and substantiate the hypothesis that NCD could benefit more from a labeled set with a large degree of semantic similarity to the unlabeled set. Specifically, we establish an extensive and large-scale benchmark with varying degrees of semantic similarity between labeled/unlabeled datasets on ImageNet by leveraging its hierarchical class structure. As a sharp contrast, the existing NCD benchmarks are developed based on labeled sets with different number of categories and images, and completely ignore the semantic relation. For (ii), we introduce a mathematical definition for quantifying the semantic similarity between labeled and unlabeled sets. In addition, we use this metric to confirm the validity of our proposed benchmark and demonstrate that it highly correlates with NCD performance. Furthermore, without quantitative analysis, previous works commonly believe that label information is always beneficial. However, counterintuitively, our experimental results show that using labels may lead to sub-optimal outcomes in low-similarity settings.

In medical genetics, each genetic variant is evaluated as an independent entity regarding its clinical importance. However, in most complex diseases, variant combinations in specific gene networks, rather than the presence of a particular single variant, predominates. In the case of complex diseases, disease status can be evaluated by considering the success level of a team of specific variants. We propose a high dimensional modelling based method to analyse all the variants in a gene network together. To evaluate our method, we selected two gene networks, mTOR and TGF-Beta. For each pathway, we generated 400 control and 400 patient group samples. mTOR and TGF-? pathways contain 31 and 93 genes of varying sizes, respectively. We produced Chaos Game Representation images for each gene sequence to obtain 2-D binary patterns. These patterns were arranged in succession, and a 3-D tensor structure was achieved for each gene network. Features for each data sample were acquired by exploiting Enhanced Multivariance Products Representation to 3-D data. Features were split as training and testing vectors. Training vectors were employed to train a Support Vector Machines classification model. We achieved more than 96% and 99% classification accuracies for mTOR and TGF-Beta networks, respectively, using a limited amount of training samples.

Technology has always had a significant impact on supply chains. In 2023, it is going to be more important than ever as the "Great Supply Chain Disruption" continues to challenge organisations around the world, according to SAPICS (The Professional Body for Supply Chain Management in Southern Africa). War, raw materials shortages, rising energy costs and extreme weather conditions are just some of the factors that will disrupt global supply chains in 2023, warns the non-profit organisation that aims to elevate, educate and empower the community of supply chain professionals across Africa. "In South Africa, the electricity crisis will continue to challenge businesses across all sectors. The negative impact on energy-intensive and irrigation dependent agricultural industries in particular will resonate through the entire supply chain – from the farm to consumers, who will have to pay more and have fewer competitive options available on supermarket shelves," says SAPICS president MJ Schoemaker.