Abstract--Dynamic graphs (DG) are often used to describe evolving interactions between nodes in real-world applications. Temporal patterns are a natural feature of DGs and are also key to representation learning. However, existing dynamic GCN models are mostly composed of static GCNs and sequence modules, which results in the separation of spatiotemporal information and cannot effectively capture complex temporal patterns in DGs. To address this problem, this study proposes a spatial-temporal graph convolutional networks with diversified transformation (STGCNDT), which includes three aspects: a) constructing a unified graph tensor convolutional network (GTCN) using tensor M-products without the need to represent spatiotemporal information separately; b) introducing three transformation schemes in GTCN to model complex temporal patterns to aggregate temporal information; and c) constructing an ensemble of diversified transformation schemes to obtain higher representation capabilities. Empirical studies on four DGs that appear in communication networks show that the proposed STGCNDT significantly outperforms state-of-the-art models in solving link weight estimation tasks due to the diversified transformations.

This paper explores the automated process of determining stem compatibility by identifying audio recordings of single instruments that blend well with a given musical context. To tackle this challenge, we present Stem-JEPA, a novel Joint-Embedding Predictive Architecture (JEPA) trained on a multi-track dataset using a self-supervised learning approach. Our model comprises two networks: an encoder and a predictor, which are jointly trained to predict the embeddings of compatible stems from the embeddings of a given context, typically a mix of several instruments. Training a model in this manner allows its use in estimating stem compatibility - retrieving, aligning, or generating a stem to match a given mix - or for downstream tasks such as genre or key estimation, as the training paradigm requires the model to learn information related to timbre, harmony, and rhythm. We evaluate our model's performance on a retrieval task on the MUSDB18 dataset, testing its ability to find the missing stem from a mix and through a subjective user study. We also show that the learned embeddings capture temporal alignment information and, finally, evaluate the representations learned by our model on several downstream tasks, highlighting that they effectively capture meaningful musical features.

The rapid advancement of AI technology has led to widespread applications of agent systems across various domains. However, the need for detailed architecture design poses significant challenges in designing and operating these systems. This paper introduces a taxonomy focused on the architectures of foundation-model-based agents, addressing critical aspects such as functional capabilities and non-functional qualities. We also discuss the operations involved in both design-time and run-time phases, providing a comprehensive view of architectural design and operational characteristics. By unifying and detailing these classifications, our taxonomy aims to improve the design of foundation-model-based agents. Additionally, the paper establishes a decision model that guides critical design and runtime decisions, offering a structured approach to enhance the development of foundation-model-based agents. Our contributions include providing a structured architecture design option and guiding the development process of foundation-model-based agents, thereby addressing current fragmentation in the field.

Large language models (LLMs) have garnered significant interest in natural language processing (NLP), particularly their remarkable performance in various downstream tasks in resource-rich languages. Recent studies have highlighted the limitations of LLMs in low-resource languages, primarily focusing on binary classification tasks and giving minimal attention to South Asian languages. These limitations are primarily attributed to constraints such as dataset scarcity, computational costs, and research gaps specific to low-resource languages. To address this gap, we present datasets for sentiment and hate speech tasks by translating from English to Bangla, Hindi, and Urdu, facilitating research in low-resource language processing. Further, we comprehensively examine zero-shot learning using multiple LLMs in English and widely spoken South Asian languages. Our findings indicate that GPT-4 consistently outperforms Llama 2 and Gemini, with English consistently demonstrating superior performance across diverse tasks compared to low-resource languages. Furthermore, our analysis reveals that natural language inference (NLI) exhibits the highest performance among the evaluated tasks, with GPT-4 demonstrating superior capabilities.

Adversarial examples add imperceptible alterations to inputs with the objective to induce misclassification in machine learning models. They have been demonstrated to pose significant challenges in domains like image classification, with results showing that an adversarially perturbed image to evade detection against one classifier is most likely transferable to other classifiers. Adversarial examples have also been studied in malware analysis. Unlike images, program binaries cannot be arbitrarily perturbed without rendering them non-functional. Due to the difficulty of crafting adversarial program binaries, there is no consensus on the transferability of adversarially perturbed programs to different detectors. In this work, we explore the robustness of malware detectors against adversarially perturbed malware. We investigate the transferability of adversarial attacks developed against one detector, against other machine learning-based malware detectors, and code similarity techniques, specifically, locality sensitive hashing-based detectors. Our analysis reveals that adversarial program binaries crafted for one detector are generally less effective against others. We also evaluate an ensemble of detectors and show that they can potentially mitigate the impact of adversarial program binaries. Finally, we demonstrate that substantial program changes made to evade detection may result in the transformation technique being identified, implying that the adversary must make minimal changes to the program binary.

This paper explores the task of automatic prediction of text spans in a legal problem description that support a legal area label. We use a corpus of problem descriptions written by laypeople in English that is annotated by practising lawyers. Inherent subjectivity exists in our task because legal area categorisation is a complex task, and lawyers often have different views on a problem, especially in the face of legally-imprecise descriptions of issues. Experiments show that training on majority-voted spans outperforms training on disaggregated ones.

Artificial Intelligence (AI) is a widely developed and adopted technology across entire industry sectors. Integrating environmental, social, and governance (ESG) considerations with AI investments is crucial for ensuring ethical and sustainable technological advancement. Particularly from an investor perspective, this integration not only mitigates risks but also enhances long-term value creation by aligning AI initiatives with broader societal goals. Yet, this area has been less explored in both academia and industry. To bridge the gap, we introduce a novel ESG-AI framework, which is developed based on insights from engagements with 28 companies and comprises three key components. The framework provides a structured approach to this integration, developed in collaboration with industry practitioners. The ESG-AI framework provides an overview of the environmental and social impacts of AI applications, helping users such as investors assess the materiality of AI use. Moreover, it enables investors to evaluate a company's commitment to responsible AI through structured engagements and thorough assessment of specific risk areas. We have publicly released the framework and toolkit in April 2024, which has received significant attention and positive feedback from the investment community. This paper details each component of the framework, demonstrating its applicability in real-world contexts and its potential to guide ethical AI investments.

Enzyme design plays a crucial role in both industrial production and biology. However, this field faces challenges due to the lack of comprehensive benchmarks and the complexity of enzyme design tasks, leading to a dearth of systematic research. Consequently, computational enzyme design is relatively overlooked within the broader protein domain and remains in its early stages. In this work, we address these challenges by introducing MetaEnzyme, a staged and unified enzyme design framework. We begin by employing a cross-modal structure-to-sequence transformation architecture, as the feature-driven starting point to obtain initial robust protein representation. Subsequently, we leverage domain adaptive techniques to generalize specific enzyme design tasks under low-resource conditions. MetaEnzyme focuses on three fundamental low-resource enzyme redesign tasks: functional design (FuncDesign), mutation design (MutDesign), and sequence generation design (SeqDesign). Through novel unified paradigm and enhanced representation capabilities, MetaEnzyme demonstrates adaptability to diverse enzyme design tasks, yielding outstanding results. Wet lab experiments further validate these findings, reinforcing the efficacy of the redesign process.

Interpersonal communication plays a key role in managing people's emotions, especially on digital platforms. Studies have shown that people use social media and consume online content to regulate their emotions and find support for rest and recovery. However, these platforms are not designed for emotion regulation, which limits their effectiveness in this regard. To address this issue, we propose an approach to enhance Interpersonal Emotion Regulation (IER) on online platforms through content recommendation. The objective is to empower users to regulate their emotions while actively or passively engaging in online platforms by crafting media content that aligns with IER strategies, particularly empathic responding. The proposed recommendation system is expected to blend system-initiated and user-initiated emotion regulation, paving the way for real-time IER practices on digital media platforms. To assess the efficacy of this approach, a mixed-method research design is used, including the analysis of text-based social media data and a user survey. Digital applications has served as facilitators in this process, given the widespread recognition of digital media applications for Digital Emotion Regulation (DER). The study collects 37.5K instances of user posts and interactions on Reddit over a year to design a Contextual Multi-Armed Bandits (CMAB) based recommendation system using features from user activity and preferences. The experimentation shows that the empathic recommendations generated by the proposed recommendation system are preferred by users over widely accepted ER strategies such as distraction and avoidance.

Enabling Large Language Models (LLMs) to generate citations in Question-Answering (QA) tasks is an emerging paradigm aimed at enhancing the verifiability of their responses when LLMs are utilizing external references to generate an answer. However, there is currently no unified framework to standardize and fairly compare different citation generation methods, leading to difficulties in reproducing different methods and a comprehensive assessment. To cope with the problems above, we introduce \name, an open-source and modular toolkit designed to facilitate the implementation and evaluation of existing citation generation methods, while also fostering the development of new approaches to improve citation quality in LLM outputs. This tool is highly extensible, allowing users to utilize 4 main modules and 14 components to construct a pipeline, evaluating an existing method or innovative designs. Our experiments with two state-of-the-art LLMs and 11 citation generation baselines demonstrate varying strengths of different modules in answer accuracy and citation quality improvement, as well as the challenge of enhancing granularity. Based on our analysis of the effectiveness of components, we propose a new method, self-RAG \snippet, obtaining a balanced answer accuracy and citation quality. Citekit is released at https://github.com/SjJ1017/Citekit.