Search

This study examines the application of adversarial attack concepts to control the evolution of cooperation in the prisoner's dilemma game in complex networks. Specifically, it proposes a simple adversarial attack method that drives players' strategies towards a target state by adding small perturbations to social networks. The proposed method is evaluated on both model and real-world networks. Numerical simulations demonstrate that the proposed method can effectively promote cooperation with significantly smaller perturbations compared to other techniques. Additionally, this study shows that adversarial attacks can also be useful in inhibiting cooperation (promoting defection). The findings reveal that adversarial attacks on social networks can be potent tools for both promoting and inhibiting cooperation, opening new possibilities for controlling cooperative behavior in social systems while also highlighting potential risks., Comment: 17 pages, 4 figures

Large Language Models (LLMs), such as ChatGPT, encounter `jailbreak' challenges, wherein safeguards are circumvented to generate ethically harmful prompts. This study introduces a straightforward black-box method for efficiently crafting jailbreak prompts, addressing the significant complexity and computational costs associated with conventional methods. Our technique iteratively transforms harmful prompts into benign expressions directly utilizing the target LLM, predicated on the hypothesis that LLMs can autonomously generate expressions that evade safeguards. Through experiments conducted with ChatGPT (GPT-3.5 and GPT-4) and Gemini-Pro, our method consistently achieved an attack success rate exceeding 80% within an average of five iterations for forbidden questions and proved robust against model updates. The jailbreak prompts generated were not only naturally-worded and succinct but also challenging to defend against. These findings suggest that the creation of effective jailbreak prompts is less complex than previously believed, underscoring the heightened risk posed by black-box jailbreak attacks., Comment: 12 pages, 4 figures, 3 tables

As large language models (LLMs) become more deeply integrated into various sectors, understanding how they make moral judgments has become crucial, particularly in the realm of autonomous driving. This study utilized the Moral Machine framework to investigate the ethical decision-making tendencies of prominent LLMs, including GPT-3.5, GPT-4, PaLM 2, and Llama 2, comparing their responses to human preferences. While LLMs' and humans' preferences such as prioritizing humans over pets and favoring saving more lives are broadly aligned, PaLM 2 and Llama 2, especially, evidence distinct deviations. Additionally, despite the qualitative similarities between the LLM and human preferences, there are significant quantitative disparities, suggesting that LLMs might lean toward more uncompromising decisions, compared to the milder inclinations of humans. These insights elucidate the ethical frameworks of LLMs and their potential implications for autonomous driving., Comment: 12 pages, 2 Figures

This study investigates adversarial attacks conducted to distort voter model dynamics in complex networks. Specifically, a simple adversarial attack method is proposed to hold the state of opinions of an individual closer to the target state in the voter model dynamics. This indicates that even when one opinion is the majority, the vote outcome can be inverted (i.e., the outcome can lean toward the other opinion) by adding extremely small (hard-to-detect) perturbations strategically generated in social networks. Adversarial attacks are relatively more effective in complex (large and dense) networks. These results indicate that opinion dynamics can be unknowingly distorted., Comment: 7 pages, 5 figures

Universal adversarial attacks, which hinder most deep neural network (DNN) tasks using only a small single perturbation called a universal adversarial perturbation (UAP), is a realistic security threat to the practical application of a DNN. In particular, such attacks cause serious problems in medical imaging. Given that computer-based systems are generally operated under a black-box condition in which only queries on inputs are allowed and outputs are accessible, the impact of UAPs seems to be limited because well-used algorithms for generating UAPs are limited to a white-box condition in which adversaries can access the model weights and loss gradients. Nevertheless, we demonstrate that UAPs are easily generatable using a relatively small dataset under black-box conditions. In particular, we propose a method for generating UAPs using a simple hill-climbing search based only on DNN outputs and demonstrate the validity of the proposed method using representative DNN-based medical image classifications. Black-box UAPs can be used to conduct both non-targeted and targeted attacks. Overall, the black-box UAPs showed high attack success rates (40% to 90%), although some of them had relatively low success rates because the method only utilizes limited information to generate UAPs. The vulnerability of black-box UAPs was observed in several model architectures. The results indicate that adversaries can also generate UAPs through a simple procedure under the black-box condition to foil or control DNN-based medical image diagnoses, and that UAPs are a more realistic security threat., Comment: 19 pages, 3 figures, 2 tables

Under the epidemic of the novel coronavirus disease 2019 (COVID-19), chest X-ray computed tomography imaging is being used for effectively screening COVID-19 patients. The development of computer-aided systems based on deep neural networks (DNNs) has been advanced, to rapidly and accurately detect COVID-19 cases, because the need for expert radiologists, who are limited in number, forms a bottleneck for the screening. However, so far, the vulnerability of DNN-based systems has been poorly evaluated, although DNNs are vulnerable to a single perturbation, called universal adversarial perturbation (UAP), which can induce DNN failure in most classification tasks. Thus, we focus on representative DNN models for detecting COVID-19 cases from chest X-ray images and evaluate their vulnerability to UAPs generated using simple iterative algorithms. We consider nontargeted UAPs, which cause a task failure resulting in an input being assigned an incorrect label, and targeted UAPs, which cause the DNN to classify an input into a specific class. The results demonstrate that the models are vulnerable to nontargeted and targeted UAPs, even in case of small UAPs. In particular, 2% norm of the UPAs to the average norm of an image in the image dataset achieves >85% and >90% success rates for the nontargeted and targeted attacks, respectively. Due to the nontargeted UAPs, the DNN models judge most chest X-ray images as COVID-19 cases. The targeted UAPs make the DNN models classify most chest X-ray images into a given target class. The results indicate that careful consideration is required in practical applications of DNNs to COVID-19 diagnosis; in particular, they emphasize the need for strategies to address security concerns. As an example, we show that iterative fine-tuning of the DNN models using UAPs improves the robustness of the DNN models against UAPs., Comment: 17 pages, 5 figures, 3 tables

Deep neural networks (DNNs) are vulnerable to adversarial attacks. In particular, a single perturbation known as the universal adversarial perturbation (UAP) can foil most classification tasks conducted by DNNs. Thus, different methods for generating UAPs are required to fully evaluate the vulnerability of DNNs. A realistic evaluation would be with cases that consider targeted attacks; wherein the generated UAP causes DNN to classify an input into a specific class. However, the development of UAPs for targeted attacks has largely fallen behind that of UAPs for non-targeted attacks. Therefore, we propose a simple iterative method to generate UAPs for targeted attacks. Our method combines the simple iterative method for generating non-targeted UAPs and the fast gradient sign method for generating a targeted adversarial perturbation for an input. We applied the proposed method to state-of-the-art DNN models for image classification and proved the existence of almost imperceptible UAPs for targeted attacks; further, we demonstrated that such UAPs are easily generatable., Comment: 4 pages, 3 figures, 1 table

Environmental contaminant exposure can pose significant risks to human health. Therefore, evaluating the impact of this exposure is of great importance; however, it is often difficult because both the molecular mechanism of disease and the mode of action of the contaminants are complex. We used network biology techniques to quantitatively assess the impact of environmental contaminants on the human interactome and diseases with a particular focus on seven major contaminant categories: persistent organic pollutants (POPs), dioxins, polycyclic aromatic hydrocarbons (PAHs), pesticides, perfluorochemicals (PFCs), metals, and pharmaceutical and personal care products (PPCPs). We integrated publicly available data on toxicogenomics, the diseasome, protein-protein interactions (PPIs), and gene essentiality and found that a few contaminants were targeted to many genes, and a few genes were targeted by many contaminants. The contaminant targets were hub proteins in the human PPI network, whereas the target proteins in most categories did not contain abundant essential proteins. Generally, contaminant targets and disease-associated proteins were closely associated with the PPI network, and the closeness of the associations depended on the disease type and chemical category. Network biology techniques were used to identify environmental contaminants with broad effects on the human interactome and contaminant-sensitive biomarkers. Moreover, this method enabled us to quantify the relationship between environmental contaminants and human diseases, which was supported by epidemiological and experimental evidence. These methods and findings have facilitated the elucidation of the complex relationship between environmental exposure and adverse health outcomes., Comment: 35 pages, 12 figures

The evolution of species habitat range is an important topic over a wide range of research fields. In higher organisms, habitat range evolution is generally associated with genetic events such as gene duplication. However, the specific factors that determine habitat variability remain unclear at higher levels of biological organization (e.g., biochemical networks). One widely accepted hypothesis developed from both theoretical and empirical analyses is that habitat variability promotes network modularity; however, this relationship has not yet been directly tested in higher organisms. Therefore, I investigated the relationship between habitat variability and metabolic network modularity using compound and enzymatic networks in flies and mammals. Contrary to expectation, there was no clear positive correlation between habitat variability and network modularity. As an exception, the network modularity increased with habitat variability in the enzymatic networks of flies. However, the observed association was likely an artifact, and the frequency of gene duplication appears to be the main factor contributing to network modularity. These findings raise the question of whether or not there is a general mechanism for habitat range expansion at a higher level (i.e., above the gene scale). This study suggests that the currently widely accepted hypothesis for habitat variability should be reconsidered., Comment: 21 pages, 4 figures

The origin of allometric scaling of metabolic rate is a long-standing question in biology. Several models have been proposed for explaining the origin; however, they have advantages and disadvantages. In particular, previous models only demonstrate either two important observations for the allometric scaling: the variability of scaling exponents and predominance of 3/4-power law. Thus, these models have a dispute over their validity. In this study, we propose a simple geometry model, and show that a hypothesis that total surface area of cells determines metabolic rate can reproduce these two observations by combining two concepts: the impact of cell sizes on metabolic rate and fractal-like (hierarchical) organization. The proposed model both theoretically and numerically demonstrates the approximately 3/4-power law although several different biological strategies are considered. The model validity is confirmed using empirical data. Furthermore, the model suggests the importance of heterogeneity of cell size for the emergence of the allometric scaling. The proposed model provides intuitive and unique insights into the origin of allometric scaling laws in biology, despite several limitations of the model., Comment: 8 pages, 5 figures

Network self-similarity or fractality are widely accepted as an important topological property of metabolic networks; however, recent studies cast doubt on the reality of self-similarity in the networks. Therefore, we perform a comprehensive evaluation of metabolic network fractality using a box-covering method with an earlier version and the latest version of metabolic networks, and demonstrate that the latest metabolic networks are almost self-dissimilar, while the earlier ones are fractal, as reported in a number of previous studies. This result may be because the networks were randomized because of an increase in network density due to database updates, suggesting that the previously observed network fractality was due to a lack of available data on metabolic reactions. This finding may not entirely discount the importance of self-similarity of metabolic networks. Rather, it highlights the need for a more suitable definition of network fractality and a more careful examination of self-similarity of metabolic networks., Comment: 7 pages, 5 figures

Ecological networks exhibit non-random structural patterns, such as modularity and nestedness, which indicate ecosystem stability, species diversity, and connectance. Such structure-stability relationships are well known. However, another important perspective is less well understood: the relationship between the environment and structure. Inspired by theoretical studies that suggest that network structure can change due to environmental variability, we collected data on a number of empirical food webs and mutualistic networks and evaluated the effect of climatic seasonality on ecological network structure. As expected, we found that climatic seasonality affects ecological network structure. In particular, an increase in modularity due to climatic seasonality was observed in food webs; however, it is debatable whether this occurs in mutualistic networks. Interestingly, the type of climatic seasonality that affects network structure differs with ecosystem type. Rainfall and temperature seasonality influence freshwater food webs and mutualistic networks, respectively; food webs are smaller, and more modular, with increasing rainfall seasonality. Mutualistic networks exhibit a higher diversity (particularly of animals) with increasing temperature seasonality. These results confirm the theoretical prediction that stability increases with greater perturbation. Although these results are still debatable because of several limitations in the data analysis, they may enhance our understanding of environment-structure relationships., Comment: 22 pages, 3 figures, 1 table

In our previous study, we showed that the branching process approximation is useful for estimating metabolic robustness, measured using the impact degree. By applying a theory of random family forests, we here extend the branching process approximation to consider the knockout of {\it multiple} reactions, inspired by the importance of multiple knockouts reported by recent computational and experimental studies. In addition, we propose a better definition of the number of offspring of each reaction node, allowing for an improved estimation of the impact degree distribution obtained as a result of a single knockout. Importantly, our proposed approach is also applicable to multiple knockouts. The comparisons between theoretical predictions and numerical results using real-world metabolic networks demonstrate the validity of the modeling based on random family forests for estimating the impact degree distributions resulting from the knockout of multiple reactions., Comment: 20 pages, 6 figures, 2 tables

Molecular signaling networks are believed to determine cancer robustness. Although cancer patient survivability was reported to correlate with the heterogeneous connectivity of the signaling networks inspired by theoretical studies on the increase of network robustness due to the heterogeneous connectivity, other theoretical and data analytic studies suggest an alternative explanation: the impact of modular organization of networks on biological robustness or adaptation to changing environments. In this study, thus, we evaluate whether the modularity--robustness hypothesis is applicable to cancer using network analysis. We focus on 14 specific cancer types whose molecular signaling networks are available in databases, and show that modular organization of cancer signaling networks is associated with the patient survival rate. In particular, the cancers with less modular signaling networks are more curable. This result is consistent with a prediction from the modularity--robustness hypothesis. Furthermore, we show that the network modularity is a better descriptor of the patient survival rate than the heterogeneous connectivity. However, these results do not contradict the importance of the heterogeneous connectivity. Rather, they provide new and different insights into the relationship between cellular networks and cancer behaviors. Despite several limitations of data analysis, these findings enhance our understanding of adaptive and evolutionary mechanisms of cancer cells., Comment: 17 pages, 5 figures

Metabolic networks consist of linked functional components, or modules. The mechanism underlying metabolic network modularity is of great interest not only to researchers of basic science but also to those in fields of engineering. Previous studies have suggested a theoretical model, which proposes that a change in the evolutionary goal (system-specific purpose) increases network modularity, and this hypothesis was supported by statistical data analysis. Nevertheless, further investigation has uncovered additional possibilities that might explain the origin of network modularity. In this work, we propose an evolving network model without tuning parameters to describe metabolic networks. We demonstrate, quantitatively, that metabolic network modularity can arise from simple growth processes, independent of the change in the evolutionary goal. Our model is applicable to a wide range of organisms, and appears to suggest that metabolic network modularity can be more simply determined than previously thought. Nonetheless, our proposition does not serve to contradict the previous model; it strives to provide an insight from a different angle in the ongoing efforts to understand metabolic evolution, with the hope of eventually achieving the synthetic engineering of metabolic networks., Comment: 11 pages, 7 figures

Theoretical frameworks to estimate the tolerance of metabolic networks to various failures are important to evaluate the robustness of biological complex systems in systems biology. In this paper, we focus on a measure for robustness in metabolic networks, namely, the impact degree, and propose an approximation method to predict the probability distribution of impact degrees from metabolic network structures using the theory of branching process. We demonstrate the relevance of this method by testing it on real-world metabolic networks. Although the approximation method possesses a few limitations, it may be a powerful tool for evaluating metabolic robustness., Comment: 17 pages, 4 figures, 4 tables

Living organisms produce metabolites of many types via their metabolisms. Especially, flavonoids, a kind of secondary metabolites, of plant species are interesting examples. Since plant species are believed to have specific flavonoids with respect to diverse environment, elucidation of design principles of metabolite distributions across plant species is important to understand metabolite diversity and plant evolution. In the previous work, we found heterogeneous connectivity in metabolite distributions, and proposed a simple model to explain a possible origin of heterogeneous connectivity. In this paper, we show further structural properties in the metabolite distribution among families inspired by analogy with plant-animal mutualistic networks: nested structure and modular structure. An earlier model represents that these structural properties in bipartite relationships are determined based on traits of elements and external factors. However, we find that the architecture of metabolite distributions is described by simple evolution processes without trait-based mechanisms by comparison between our model and the earlier model. Our model can better predict nested structure and modular structure in addition to heterogeneous connectivity both qualitatively and quantitatively. This finding implies an alternative possible origin of these structural properties, and suggests simpler formation mechanisms of metabolite distributions across plant species than expected., Comment: 14 pages, 5 figures

Nested structure, which is non-random, controls cooperation dynamics and biodiversity in plant-animal mutualistic networks. This structural pattern has been explained in a static (non-growth) network models. However, evolutionary processes might also influence the formation of such a structural pattern. We thereby propose an evolving network model for plant-animal interactions and show that non-random patterns such as nested structure and heterogeneous connectivity are both qualitatively and quantitatively predicted through simple evolutionary processes. This finding implies that network models can be simplified by considering evolutionary processes, and also that another explanation exists for the emergence of non-random patterns and might provide more comprehensible insights into the formation of plant-animal mutualistic networks from the evolutionary perspective., Comment: 12 pages, 4 figures

We investigate the distribution of flavonoids, a major category of plant secondary metabolites, across species. Flavonoids are known to show high species specificity, and were once considered as chemical markers for understanding adaptive evolution and characterization of living organisms. We investigate the distribution among species using bipartite networks, and find that two heterogeneous distributions are conserved among several families: the power-law distributions of the number of flavonoids in a species and the number of shared species of a particular flavonoid. In order to explain the possible origin of the heterogeneity, we propose a simple model with, essentially, a single parameter. As a result, we show that two respective power-law statistics emerge from simple evolutionary mechanisms based on a multiplicative process. These findings provide insights into the evolution of metabolite diversity and characterization of living organisms that defy genome sequence analysis for different reasons., Comment: 16 pages, 7 figures, 1 table

We demonstrate the advantages of feedforward loops using a Boolean network, which is one of the discrete dynamical models for transcriptional regulatory networks. After comparing the dynamical behaviors of network embedded feedback and feedforward loops, we found that feedforward loops can provide higher temporal order (coherence) with lower entropy (randomness) in a temporal program of gene expression. In addition, complexity of the state space that increases with longer length of attractors and greater number of attractors is also reduced for networks with more feedforward loops. Feedback loops show opposite effects on dynamics of the networks. These results suggest that feedforward loops are one of the favorable local structures in biomolecular and neuronal networks., Comment: 9 Figures and 2 Tables, accepted in AROB'08. accepted in AROB'08

We propose the n-clique network as a powerful tool for understanding global structures of combined highly-interconnected subgraphs, and provide theoretical predictions for statistical properties of the n-clique networks embedded in a complex network using the degree distribution and the clustering spectrum. Furthermore, using our theoretical predictions, we find that the statistical properties are invariant between 3-clique networks and original networks for several observable real-world networks with the scale-free connectivity and the hierarchical modularity. The result implies that structural properties are identical between the 3-clique networks and the original networks., Comment: 12 pages, 5 figures

We performed a numerical study on random Boolean networks with power-law rank outdegree distributions to find local structural cause for the emergence of high or low degree of coherence in binary state variables. The degree of randomness and coherence of the binary sequence are measured by entropy and mutual information, depending on local structure that consists of a node with a highly connected, called hub and its upstream nodes, and types of Boolean functions for the nodes. With a large number of output connections from a hub, the effects of Boolean function on the hub are more prominent. The local structures that give larger entropy tend to give rise to larger mutual information. Based on the numerical results and structural conditions we derived a time-independent transmission characteristic function of state variables for the local structures. We obtained good relationships between the numerical and analytical results, which indicate that dynamical properties from the whole networks can be inferred from the differences in the local structures., Comment: 6 pages, 6 figures, 3 tables. submitted to AROB 12th '07

We propose a growing network model that consists of two tunable mechanisms: growth by merging modules which are represented as complete graphs and a fitness-driven preferential attachment. Our model exhibits the three prominent statistical properties are widely shared in real biological networks, for example gene regulatory, protein-protein interaction, and metabolic networks. They retain three power law relationships, such as the power laws of degree distribution, clustering spectrum, and degree-degree correlation corresponding to scale-free connectivity, hierarchical modularity, and disassortativity, respectively. After making comparisons of these properties between model networks and biological networks, we confirmed that our model has inference potential for evolutionary processes of biological networks., Comment: 19 pages, 8 figures

We propose a model for evolving networks by merging building blocks represented as complete graphs, reminiscent of modules in biological system or communities in sociology. The model shows power-law degree distributions, power-law clustering spectra and high average clustering coefficients independent of network size. The analytical solutions indicate that a degree exponent is determined by the ratio of the number of merging nodes to that of all nodes in the blocks, demonstrating that the exponent is tunable, and are also applicable when the blocks are classical networks such as Erd\H{o}s-R\'enyi or regular graphs. Our model becomes the same model as the Barab\'asi-Albert model under a specific condition., Comment: 8 pages, 8 figures