Nowadays, regulatory compliance has become a cornerstone of corporate governance, ensuring adherence to systematic legal frameworks. At its core, financial regulations often comprise highly intricate provisions, layered logical structures, and numerous exceptions, which inevitably result in labor-intensive or comprehension challenges. To mitigate this, recent Regulatory Technology (RegTech) and Large Language Models (LLMs) have gained significant attention in automating the conversion of regulatory text into executable compliance logic. However, their performance remains suboptimal particularly when applied to Chinese-language financial regulations, due to three key limitations: (1) incomplete domain-specific knowledge representation, (2) insufficient hierarchical reasoning capabilities, and (3) failure to maintain temporal and logical coherence. One promising solution is to develop a domain specific and code-oriented datasets for model training. Existing datasets such as LexGLUE, LegalBench, and CODE-ACCORD are often English-focused, domain-mismatched, or lack fine-grained granularity for compliance code generation. To fill these gaps, we present Compliance-to-Code, the first large-scale Chinese dataset dedicated to financial regulatory compliance. Covering 1,159 annotated clauses from 361 regulations across ten categories, each clause is modularly structured with four logical elements-subject, condition, constraint, and contextual information-along with regulation relations. We provide deterministic Python code mappings, detailed code reasoning, and code explanations to facilitate automated auditing. To demonstrate utility, we present FinCheck: a pipeline for regulation structuring, code generation, and report generation.

Rule-following agents tasked with executing policies and regulations often fail via Silent Scope Omission (SSO): a model applies a general rule but silently drops nested exceptions or counter-exceptions, producing outputs that appear compliant yet break on important edge cases. Although such failures are often framed as an agentic-systems problem, the underlying bottleneck is statutory and policy understanding, a capability typically studied in legal NLP. However, most existing legal NLP benchmarks emphasize end-task outcomes, which can overlook the structural omissions that cause SSO. To diagnose and mitigate SSO, we introduce NormBench, a benchmark of 2,290 provisions spanning Chinese (laws and local policies), English (U.S. tax law, GDPR, and corporate policies), and cross-lingual settings, designed for defeasible scope parsing: identifying precisely which clause overrides which. NormBench uses Span-Grounded Deontic Trees (SG-DT), a compiler-style intermediate representation that anchors every logical branch to source spans and requires explicit exclusion guards, enabling deterministic compilation and audit. Evaluations of frontier LLMs reveal two recurring pathologies: (1) Recursion Decay, where performance drops sharply as defeater depth increases, and (2) an Auditability Trap, where models retrieve relevant spans but fail to assemble correct control flow. Using SG-DT as a constrained intermediate output improves whole-tree fidelity and defeater recovery, and downstream experiments show that its utility is mechanism-specific: gains concentrate on exception-active, SSO-prone cases, while aggregate accuracy can be mixed when the added structure is unnecessary or parser fidelity is low.

Generative artificial intelligence (AI) exhibits the capability to generate creative content akin to human output with greater efficiency and reduced costs. This groundbreaking capability, however, has ignited a debate regarding its potential to displace human creators. In light of these discussions, this paper empirically investigates the impact of generative AI on market equilibrium, in the context of China’s leading art outsourcing platform. We overcome the challenge of causal inference by identifying an unanticipated and sudden leak of an advanced image-generative AI as a natural experiment. This leak precipitated a notable reduction in the production costs of anime-style images compared to other genres, thereby providing a unique opportunity for difference-in-differences comparisons. Our analysis shows that the advent of generative AI led to a 64

Natural Language to SQL (NL2SQL) provides a new model-centric paradigm that simplifies database access for non-technical users by converting natural language queries into SQL commands. Recent advancements, particularly those integrating Retrieval-Augmented Generation (RAG) and Chain-of-Thought (CoT) reasoning, have made significant strides in enhancing NL2SQL performance. However, challenges such as inaccurate task decomposition and keyword extraction by LLMs remain major bottlenecks, often leading to errors in SQL generation. While existing datasets aim to mitigate these issues by fine-tuning models, they struggle with over-fragmentation of tasks and lack of domain-specific keyword annotations, limiting their effectiveness.To address these limitations, we present DeKeyNLU, a novel dataset which contains 1,500 meticulously annotated QA pairs aimed at refining task decomposition and enhancing keyword extraction precision for the RAG pipeline. Fine-tuned with DeKeyNLU, we propose DeKeySQL, a RAG-based NL2SQL pipeline that employs three distinct modules for user question understanding, entity retrieval, and generation to improve SQL generation accuracy. We benchmarked multiple model configurations within DeKeySQL RAG pipeline. Experimental results demonstrate that fine-tuning with DeKeyNLU significantly improves SQL generation accuracy on both BIRD (62.31% to 69.10%) and Spider (84.2% to 88.7%) dev datasets.

Questionnaire-based surveys are foundational to social science research and public policymaking, yet traditional survey methods remain costly, time-consuming, and often limited in scale. This paper explores a new paradigm: simulating virtual survey respondents using Large Language Models (LLMs). We introduce two novel simulation settings, namely Partial Attribute Simulation (PAS) and Full Attribute Simulation (FAS), to systematically evaluate the ability of LLMs to generate accurate and demographically coherent responses. In PAS, the model predicts missing attributes based on partial respondent profiles, whereas FAS involves generating complete synthetic datasets under both zero-context and context-enhanced conditions. We curate a comprehensive benchmark suite, LLM-S^3 (Large Language Model-based Sociodemographic Survey Simulation), that spans 11 real-world public datasets across four sociological domains. Our evaluation of multiple mainstream LLMs (GPT-3.5/4 Turbo, LLaMA 3.0/3.1-8B) reveals consistent trends in prediction performance, highlights failure modes, and demonstrates how context and prompt design impact simulation fidelity. This work establishes a rigorous foundation for LLM-driven survey simulations, offering scalable and cost-effective tools for sociological research and policy evaluation. Our code and dataset are available at: https://github.com/dart-lab-research/LLM-S-Cube-Benchmark

Meteorological heatmaps play a vital role in deciphering extreme weather phenomena, yet their inherent complexities marked by irregular contours, unstructured patterns, and complex color variations present unique analytical hurdles for state-of-the-art Vision-Language Models (VLMs). Current state-of-the-art models like GPT-4o, Qwen-VL, and LLaVA 1.6 struggle with tasks such as precise color identification and spatial localization, resulting in inaccurate or incomplete interpretations. To address these challenges, we introduce Sparse Position and Outline Tracking (SPOT), a novel algorithm specifically designed to process irregularly shaped colored regions in visual data. SPOT identifies and localizes these regions by extracting their spatial coordinates, enabling structured representations of irregular shapes. Building on SPOT, we construct ClimateIQA, a novel meteorological visual question answering (VQA) dataset, comprising 26,280 high-resolution heatmaps and 762,120 instruction samples for wind gust, total precipitation, wind chill index and heat index analysis. ClimateIQA enhances VLM training by incorporating spatial cues, geographic metadata, and reanalysis data, improving model accuracy in interpreting and describing extreme weather features. Furthermore, we develop Climate-Zoo, a suite of fine-tuned VLMs based on SPOT-empowered ClimateIQA, which significantly outperforms existing models in meteorological heatmap tasks.

Hierarchical Multi-Label Classification (HMLC) is a well-established problem that aims at assigning data instances to multiple classes stored in a hierarchical structure. Despite its importance, existing approaches often face two key limitations: (i) They employ dense networks to solely explore the class hierarchy as hard criterion for maintaining taxonomic consistency among predicted classes, yet without leveraging rich semantic relationships between instances and classes; (ii) They struggle to generalize in settings with deep class levels, since the mini-batches uniformly sampled from different levels ignore the varying complexities of data and result in a non-smooth model adaptation to sparse data. To mitigate these issues, we present a Self-Paced Unified Representation (SPUR) learning framework, which focuses on the interplay between instance and classes to flexibly organize the training process of HMLC algorithms. Our framework consists of two lightweight encoders designed to capture the semantics of input features and the topological information of the class hierarchy. These encoders generate unified embeddings of instances and class hierarchy, which enable SPUR to exploit semantic dependencies between them and produce predictions in line with taxonomic constraints. Furthermore, we introduce a dynamic hardness measurement strategy that considers both class hierarchy and instance features to estimate the learning difficulty of each instance. This strategy is achieved by incorporating the propagation loss obtained at each hierarchical level, allowing for a more comprehensive assessment of learning complexity. Extensive experiments on several empirical benchmarks demonstrate the effectiveness and efficiency of SPUR compared to state-of-the-art methods, especially in scenarios with missing features.

In recent years, Graph Neural Networks (GNNs) and Large Language Models (LLMs) have exhibited remarkable capability in addressing different graph learning and natural language tasks, respectively. Motivated by this, integrating LLMs with GNNs has been increasingly studied to acquire transferable knowledge across modalities, which leads to improved empirical performance in language and graph domains. However, existing studies mainly focused on a single-domain scenario by designing complicated integration techniques to manage multimodal data effectively. Therefore, a concise and generic learning framework for multi-domain tasks, i.e., graph and language domains, is highly desired yet remains under-exploited due to two major challenges. First, the language corpus of downstream tasks differs significantly from graph data, making it hard to bridge the knowledge gap between modalities. Second, not all knowledge demonstrates immediate benefits for downstream tasks, potentially introducing disruptive noise to context-sensitive models like LLMs. To tackle these challenges, we propose a novel plug-and-play framework for incorporating a lightweight cross-domain prompting method into both language and graph learning tasks. Specifically, we first convert the textual input into a domain-scalable prompt, which not only preserves the semantic and logical contents of the textual input, but also highlights related graph information as external knowledge for different domains. Then, we develop a reinforcement learning-based method to learn the optimal edge selection strategy for useful knowledge extraction, which profoundly sharpens the multi-domain model capabilities. In addition, we introduce a joint multi-view optimization module to regularize agent-level collaborative learning across two domains. Finally, extensive empirical justifications over 23 public and synthetic datasets demonstrate that our approach can be applied to diverse multi-domain tasks more accurately, robustly, and reasonably, and improve the performances of the state-of-the-art graph and language models in different learning paradigms.

In recent years, Optimized Cost Per Click (OCPC) and Optimized Cost Per Mille (OCPM) have emerged as the most widely adopted pricing models in the online advertising industry. However, the existing literature has yet to identify the specific conditions under which these models outperform traditional pricing models like Cost Per Click (CPC) and Cost Per Action (CPA). To fill the gap, this paper builds an economic model that compares OCPC with CPC and CPA theoretically, which incorporates out-site scenarios and outside options as two key factors. Our analysis reveals that OCPC can effectively replace CPA by tackling the problem of advertisers strategically manipulating conversion reporting in out-site scenarios where conversions occur outside the advertising platform. Furthermore, OCPC exhibits the potential to surpass CPC in platform payoffs by providing higher advertiser payoffs and consequently attracting more advertisers. However, if advertisers have less competitive outside options and consistently stay in the focal platform, the platform may achieve higher payoffs using CPC. Our findings deliver valuable insights for online advertising platforms in selecting optimal pricing models, and provide recommendations for further enhancing their payoffs. To the best of our knowledge, this is the first study to analyze OCPC from an economic perspective. Moreover, our analysis can be applied to the OCPM model as well.

Integrating data-driven and rule-based approaches is crucial for therapy recommendations since they can collaborate to achieve better performance. Medical rules, which are chains of reasoning that can infer therapies, widely exist. However, their symbolic and logical forms make integrating them with data-driven modeling technologies hard. Although rare attempts have indirectly modeled rules using data that supports them, the poor generalization of medical rules leads to inadequate supporting data and thus impairs the benefit of medical rules. To this end, we propose R2V-MIF, which fills the gap by rule-to-vector contrastive learning (R2V) and multi-channel information fusion (MIF). R2V is a data-free module and utilizes a hypergraph, including condition and result nodes, to instantiate the logic of medical rules. Each rule is reflected in the relations between nodes, and their representations are determined through contrastive learning. By taking rule representations as a bridge, MIF integrates the knowledge from medical rules, similar neighbors, and patient contents, and then recommends therapies. Extensive experiments show that R2V-MIF outperforms the baselines in several metrics using real-world medical data. Our code is available at https://github.com/vgeek-z/r2vmif.

Recent years have witnessed the fast development of deep learning techniques in quantitative trading. It still remains unclear how to exploit deep learning techniques to improve high-frequency trading (HFT). Indeed, there are two emerging challenges for the use of deep learning for HFT: (i) how to quantify fast-changing market conditions for tick-level signal prediction; (ii) how to establish a unified trading paradigm for different securities of diverse market conditions and severe signal sparsity. To this end, in this paper, we propose an Online Adaptive Sequence Learning (LEVER) framework, which consists of two distinct components to predict the HFT signals at the tick level for a variety of securities simultaneously. Specifically, we start with a single learner that adopts an encoder-decoder architecture for each security-based HFT signal prediction. In this single learner, an ordered encoder module first captures the variability patterns of the security's price curve by encoding the input indicator sequence from different time ranges. An unordered decoder module then outlines the pivot points of the price curve as support and resistance levels to quantify the market status. Based on the measured market condition, a prediction module further approximates the impacts of upcoming security data as the potential market momentum to detect the tick-level trading signals. To overcome the computational challenges and signal sparsity posed by online HFT for multiple securities, we develop a competitive active-meta learning paradigm to enhance the signal learners' learning efficiency for online implementation. Finally, extensive experiments on real-world stock market data demonstrate the effectiveness of our deployed LEVER for improving the performances of the existing industry method by 0.27 in the Sharpe ratio and by 0.09% in a transaction-based return.

Venture capital investments play a powerful role in fueling the emergence and growth of early-stage startups. However, only a small fraction of venture-backed startups can survive and exit successfully. Prior data-driven prediction based or recommendation based solutions are incapable of providing effective and actionable strategies on proper investment timing and amounts for startups across different investment rounds. In this paper, we develop a novel reinforcement learning-based method, AlphaVC, to facilitate venture capitalists' decision-making. Our policy-based reinforcement learning agents can dynamically identify the best candidates and sequentially place the optimal investment amounts at proper rounds to maximize financial returns for a given portfolio. We retrieve company demographics and investment activity data from Crunchbase. Our methodology demonstrates its efficacy and superiority in both ranking and portfolio-based performance metrics in comparison with various state-of-the-art baseline methods. Through sensitivity and ablation analyses, our research highlights the significance of factoring in the distal outcome and acknowledging the learning effect when making decisions at different time points. Additionally, we observe that AlphaVC concentrates on a select number of high-potential companies, but distributes investments evenly across various stages of the investment process.

As a key component of e-commerce computing, product representation learning (PRL) provides benefits for a variety of applications, including product matching, search, and categorization. The existing PRL approaches have poor language understanding ability due to their inability to capture contextualized semantics. In addition, the learned representations by existing methods are not easily transferable to new products. Inspired by the recent advance of pre-trained language models (PLMs), we make the attempt to adapt PLMs for PRL to mitigate the above issues. In this article, we develop KINDLE, a Knowledge-drIven pre-trainiNg framework for proDuct representation LEarning, which can preserve the contextual semantics and multi-faceted product knowledge robustly and flexibly. Specifically, we first extend traditional one-stage pre-training to a two-stage pre-training framework, and exploit a deliberate knowledge encoder to ensure a smooth knowledge fusion into PLM. In addition, we propose a multi-objective heterogeneous embedding method to represent thousands of knowledge elements. This helps KINDLE calibrate knowledge noise and sparsity automatically by replacing isolated classes as training targets in knowledge acquisition tasks. Furthermore, an input-aware gating network is proposed to select the most relevant knowledge for different downstream tasks. Finally, extensive experiments have demonstrated the advantages of KINDLE over the state-of-the-art baselines across three downstream tasks.

Earnings calls are periodically held to disclose the past performance and future plan of a publicly-traded company. The analysis of earnings call transcripts has come under the spotlight of both Artificial Intelligence (AI) and finance communities due to the critical value for understanding corporate fundamentals. However, existing studies have been limited by the challenges inherent in the lengthy, noisy, and scarce transcripts, and thus are difficult in learning to distill salient market-influencing semantics from these transcripts. To this end, in this paper, we propose a text encoder with a sparse self-attention mechanism to address the lengthiness and noisiness issues. Also, a multi-source counterfactual augmentation framework is developed to leverage abundant unlabeled texts (e.g., financial news) to address the data scarcity issue. In particular, the unlabeled texts serve as external perturbations to the original transcripts under strong semantic and task-specific constraints, which helps for efficient model learning. Finally, extensive experiments on real-world financial data demonstrate that our approach can effectively perform both short-term and long-term volatility prediction as well as financial sentiment analysis. Moreover, we showcase how the proposed framework can enhance model interpretability and provide high-quality augmented data.

The fate of unimported mitochondrial precursors has been increasingly studied in recent years, mostly focusing on protein degradation. In this issue of the EMBO journal, Krämer et al discovered MitoStores, a new protective mechanism that temporarily stores mitochondrial proteins in cytosolic deposits.

Many real-world Internet of Things (IoT) systems, which include a variety of Internet-connected sensory devices, produce substantial amounts of multivariate time-series data. Meanwhile, vital IoT infrastructures, such as smart power grids and water distribution networks are frequently targeted by cyberattacks, making anomaly detection an important study topic. Modeling such relatedness is, nevertheless, unavoidable for any efficient and effective anomaly detection system, given the intricate topological and nonlinear connections that are originally unknown among sensors. Furthermore, detecting anomalies in multivariate time series is difficult due to their temporal dependency and stochasticity. This article presented GTA, a new framework for multivariate time-series anomaly detection that involves automatically learning a graph structure, graph convolution, and modeling temporal dependency using a transformer-based architecture. The connection learning policy, which is based on the Gumbel-softmax sampling approach to learn bidirected links among sensors directly, is at the heart of the learning graph structure. To describe the anomaly information flow between network nodes, we introduced a new graph convolution called influence propagation convolution. In addition, to tackle the quadratic complexity barrier, we suggested a multibranch attention mechanism to replace the original multihead self-attention method. Extensive experiments on four publicly available anomaly detection benchmarks further demonstrate the superiority of our approach over alternative state of the arts.

Graph walking based on reinforcement learning (RL) has shown great success in navigating an agent to automatically complete various reasoning tasks over an incomplete knowledge graph (KG) by exploring multi-hop relational paths. However, existing multi-hop reasoning approaches only work well on short reasoning paths and tend to miss the target entity with the increasing path length. This is undesirable for many reasoning tasks in real-world scenarios, where short paths connecting the source and target entities are not available in incomplete KGs, and thus the reasoning performances drop drastically unless the agent is able to seek out more clues from longer paths. To address the above challenge, in this paper, we propose a dual-agent reinforcement learning framework, which trains two agents (Giant and Dwarf) to walk over a KG jointly and search for the answer collaboratively. Our approach tackles the reasoning challenge in long paths by assigning one of the agents (Giant) searching on cluster-level paths quickly and providing stage-wise hints for another agent (Dwarf). Finally, experimental results on several KG reasoning benchmarks show that our approach can search answers more accurately and efficiently, and outperforms existing RL-based methods for long path queries by a large margin.

Earnings call (EC), as a periodic teleconference of a publicly-traded company, has been extensively studied as an essential market indicator because of its high analytical value in corporate fundamentals. The recent emergence of deep learning techniques has shown great promise in creating automated pipelines to benefit the EC-supported financial applications. However, these methods presume all included contents to be informative without refining valuable semantics from long-text transcript and suffer from EC scarcity issue. Meanwhile, these black-box methods possess inherent difficulties in providing human-understandable explanations. To this end, in this paper, we propose a Multi-Domain Transformer-Based Counterfactual Augmentation, named MTCA, to address the above problems. Specifically, we first propose a transformer-based EC encoder to attentively quantify the task-inspired significance of critical EC content for market inference. Then, a multi-domain counterfactual learning framework is developed to evaluate the gradient-based variations after we perturb limited EC informative texts with plentiful cross-domain documents, enabling MTCA to perform unsupervised data augmentation. As a bonus, we discover a way to use non-training data as instance-based explanations for which we show the result with case studies. Extensive experiments on the real-world financial datasets demonstrate the effectiveness of interpretable MTCA for improving the volatility evaluation ability of the state-of-the-art by 14.2\% in accuracy.