We present AgentSeer, an interactive observability framework for agentic AI systems. Unlike conventional tracing tools that expose raw spans or model-centric metrics, AgentSeer introduces a dual graph decomposition constructed through a deterministic rule-based parser: a temporal action graph, where each prompt or tool invocation is represented as a distinct action, and a component graph capturing architectural relations among agents, tools, and memory modules. Beyond visualization, AgentSeer enables action-level red teaming, where jailbreak payloads are systematically attached to every action node (including agent messages, tool calls, and memory retrievals) to uncover vulnerabilities invisible to model-level testing. Our demonstration features a six-agent hierarchical testbed with interactive visualization and deployment-oriented safety evaluation applied directly on the same prompts and contexts, systematically revealing high-risk interactions, context-dependent vulnerabilities, and emergent behaviors. By combining structured decomposition, automated red teaming, and rule-based reliability, AgentSeer establishes a safety-first methodology for observability in multi-agent AI.

The deployment of large language models (LLMs) raises concerns regarding their cultural misalignment and potential ramifications on individuals and societies with diverse cultural backgrounds. While the discourse has focused mainly on political and social biases, our research proposes a Cultural Alignment Test (Hoftede’s CAT) to quantify cultural alignment using Hofstede’s cultural dimension framework, which offers an explanatory cross-cultural comparison through the latent variable analysis. We apply our approach to quantitatively evaluate LLMs—namely Llama 2, GPT-3.5, and GPT-4—against the cultural dimensions of regions like the United States, China, and Arab countries, using different prompting styles and exploring the effects of language-specific fine-tuning on the models’ behavioural tendencies and cultural values. Our results quantify the cultural alignment of LLMs and reveal the difference between LLMs in explanatory cultural dimensions. Our study demonstrates that while all LLMs struggle to grasp cultural values, GPT-4 shows a unique capability to adapt to cultural nuances, particularly in Chinese settings. However, it faces challenges with American and Arab cultures. The research also highlights that fine-tuning LLama 2 models with different languages changes their responses to cultural questions, emphasizing the need for culturally diverse development in AI for worldwide acceptance and ethical use. For more details or to contribute to this research, visit our GitHub page https://github.com/reemim/Hofstedes_CAT

This paper investigates a series of machine learning models (e.g. ANN, LSTM, SVR) to predict gold prices according to traditional indices, emerging indicators, commodities, and historical price time series of gold. In our approach, three machine learning algorithms, Artificial Neural Network (ANN), Long Short-Term Memory (LSTM), and Support Vector Regression (SVR), are applied to build the models that forecast the gold price. The dataset for this research is a time-series from 1st January 2017 to 31st December 2020, containing two major indices in the US (S&P 500 and DJI), two popular cryptocurrencies (BTC and ETH), two commodities (silver and crude oil), USD index (United States Dollar against Euro), and the gold prices (historical price and volatility) [24]. The evaluation benchmarks are Mean Absolute Error (MAE), Root Mean Square Error (RMSE), and Mean Absolute Percentage Error (MAPE). In the first stage, a comparative analysis is applied to three models. In the second stage, the assessment of the impact of cryptocurrency on the models is demonstrated. It was observed that the SVR model outperforms the other two models, and our result indicates that the additional data of cryptocurrencies has a positive impact on all three models.

The data science technologies of artificial intelligence (AI), Internet of Things (IoT), big data and behavioral/predictive analytics, and blockchain are poised to revolutionize government and create a new generation of GovTech start-ups.The impact from the 'smartification' of public services and the national infrastructure will be much more significant in comparison to any other sector given government's function and importance to every institution and individual.Potential GovTech systems include Chatbots and intelligent assistants for public engagement, Robo-advisors to support civil servants, real-time management of the national infrastructure using IoT and blockchain, automated compliance/regulation, public records securely stored in blockchain distributed ledgers, online judicial and dispute resolution systems, and laws/statutes encoded as blockchain smart contracts.Government is potentially the major 'client' and also 'public champion' for these new data technologies.This review paper uses our simple taxonomy of government services to provide an overview of data science automation being deployed by governments world-wide.The goal of this review paper is to encourage the Computer Science community to engage with government to develop these new systems to transform public services and support the work of civil servants.

Derivative traders are usually required to scan through hundreds, even thousands of possible trades on a daily-basis; a concrete case is the so-called Mid-Curve Calendar Spread (MCCS). The actual procedure in place is full of pitfalls and a more systematic approach where more information at hand is crossed and aggregated to find good trading picks can be highly useful and undoubtedly increase the trader's productivity. Therefore, in this work we propose an MCCS Recommendation System based on a stacking approach through Neural Networks. In order to suggest that such approach is methodologically and computationally feasible, we used a list of 15 different types of US Dollar MCCSs regarding expiration, forward and swap tenure. For each MCCS, we used 10 years of historical data ranging weekly from Sep/06 to Sep/16. Then, we started the modelling stage by: (i) fitting the base learners using as the input sensitivity metrics linked with the MCCS at time t, and its subsequent annualized returns as the output; (ii) feeding the prediction from each base model to a particular stacker; and (iii) making predictions and comparing different modelling methodologies by a set of performance metrics and benchmarks. After establishing a backtesting engine and setting performance metrics, our results suggest that our proposed Neural Network stacker compared favourably to other combination procedures.

Legal Services industries are entering a period of major disruption caused by new legal technologies (LawTech), such as artificial intelligence (AI), Internet of Things (IoT) and Blockchain. An area already undergoing major innovation is alternative dispute resolution (Alternative Dispute Resolution (2017) Wikipedia, https://en.wikipedia.org/wiki/Alternative_dispute_resolution), especially automated online dispute resolution (Online Dispute Resolution (2017) Wikipedia, https://en.wikipedia.org/wiki/Online_dispute_resolution; Katsh, E. and Rifkin, J. (2001) Online Dispute Resolution: Resolving Conflicts in Cyberspace. Jossey-Bass Wiley, New Jersey). In terms of LawTech, we broadly divide online dispute resolution into (a) Consumer ODR—uses technology to facilitate the resolution of disputes between ecommerce parties, typically online suppliers and consumers; (b) Judicial ODR—covers any means of settling ‘ordinary’ disputes where there is a hearing (using technology) but outside of the courtroom, such as divorce or personal injury cases; and what we refer to as (c) Corporate ODR—the use of technology to manage the resolution of any contractual disputes that may emerge from major multi-partner projects or financial transactions. This paper focuses on ODR and specifically the future use of automating anticipatory Corporate dispute resolution using AI and blockchain technologies. The paper describes the legal sector, and how it is being radically transformed by computer science.

Neural Network methods for stock prediction can be used to successfully signal when to buy individual stocks. The formation of weighted portfolios of such signaled stocks has however received little attention in the literature. Classical Mean-Variance based portfolio optimization techniques assume that stock returns fall into the Elliptical Family of Distributions and as such are not well suited to use with Neural Network based predictors. This paper introduces a new distribution independent framework for stock portfolio construction. Testing shows that the framework could be used to form profitable stocks portfolios when applied to a Neural Network stock predictor.

The systematic trading of equities forms the basis of the asset management industry. Analysts are trying to outperform a passive investment in an index such as the S&P 500 Index. However, statistics have shown that most analysts fail to consistently beat the index. A number of Neural Network based methods for detecting trading opportunities on Futures contracts on the S&P 500 Index have been published in the literature. However, such methods have generally been unable to demonstrate sustained performance over a significant period of time. The authors of this paper show, through the application of over ten years of experience in quantitative modelling and trading, a different type of Neural Network approach to beating the S&P 500 Index. Rather than trading Futures contracts, it is shown that by using Neural Networks to intelligently select just a handful of stocks a performance significantly in excess of a buy and hold position on the S&P 500 Index could have been achieved over a seven year period. The effect of transaction costs is also considered.

While the prediction accuracy of a large-scale recommender system can generally be improved by learning from more and more training data over time, it is unclear how well a fixed predictive model can handle the changing business dynamics in a real-world scenario. The adjustment of a predictive model is controlled by the hyperparameter settings of a selected algorithm. Although the problem of hyperparameter optimization has been studied for decades in various disciplines, the adaptiveness of the initially selected model is not as well understood. This paper presents an approach to continuously re-select hyperparameter settings of the algorithm in a large-scale retail recommender system. In particular, an automatic hyperparameter optimization technique is applied on collaborative filtering algorithms in order to improve prediction accuracy. Experiments have been conducted on a large-scale real retail dataset to challenge traditional assumption that a one-off initial hyperparameter optimization is sufficient. The proposed approach has been compared with a baseline approach and a widely used approach with two scalable collaborative filtering algorithms. The evaluations of our experiments are based on a 2-year real purchase transaction dataset of a large retail chain business, both its online e-commerce site and its offline retail stores. It is demonstrated that continuous hyperparameter optimization can effectively improve the prediction accuracy of a recommender system. This paper presents a new direction in improving the prediction performance of a large-scale recommender system.

Four major areas of research are involved in attempting to identify the fifth generation of computers (cir. 1990). The investigation of: knowledge processing systems, data and demand driven computers, integrating communications and computers, and VLSI processor architectures. From these four areas, two approaches for the fifth generation are emerging: one “revolutionary” - a parallel logic machine supporting knowledge processing applications, and the other “evolutionary” - a decentralised control flow system consisting of a network of heterogeneous processors. This paper describes the above four areas of research and discusses how their computing technologies are converging to produce fifth generation computers. It then contrasts the revolutionary logic machine approach adopted by Japan's Fifth Generation Project, and favoured by the artificial intelligence community, with the evolutionary control flow computer approach favoured by the data communications and microelectronics communities.

We present a general-purpose computer architecture based on the concept of recursion, suitable for VLSI computer systems built from replicated (LEGO-like) computing elements. The recursive computer architecture is defined by presenting a program organisation, a machine organisation and an experimental machine implementation oriented to VLSI. The program organisation, called recursive control flow, attempts a synthesis of the concepts underlying traditional control flow, data flow and reduction, to exploit the individual strengths of each organisation. The machine organisation is based on replicated general-purpose computing elements, as well as special-purpose computing elements that allow the function of individual computer systems to be specialised. These elements are interconnected to form a larger computer system and cooperate in the concurrent execution of a program. The experimental implementation is being restricted to simple, identical microcomputers each containing a memory, a processor and a communications capability. This future generation of LEGO-like computer systems are termed fifth generation computers [1] by the Japanese.

The design of a multi-processor reduction machine, supporting a class of user-defined reduction languages, is presented.