Seyed Amir Ahmad Safavi-Naini, Elahe Meftah, Josh Mohess
et al.
The competency of any intelligent agent is bounded by its formal account of the world in which it operates. Clinical AI lacks such an account. Existing frameworks address evaluation, regulation, or system design in isolation, without a shared model of the clinical world to connect them. We introduce the Clinical World Model, a framework that formalizes care as a tripartite interaction among Patient, Provider, and Ecosystem. To formalize how any agent, whether human or artificial, transforms information into clinical action, we develop parallel decision-making architectures for providers, patients, and AI agents, grounded in validated principles of clinical cognition. The Clinical AI Skill-Mix operationalizes competency through eight dimensions. Five define the clinical competency space (condition, phase, care setting, provider role, and task) and three specify how AI engages human reasoning (assigned authority, agent facing, and anchoring layer). The combinatorial product of these dimensions yields a space of billions of distinct competency coordinates. A central structural implication is that validation within one coordinate provides minimal evidence for performance in another, rendering the competency space irreducible. The framework supplies a common grammar through which clinical AI can be specified, evaluated, and bounded across stakeholders. By making this structure explicit, the Clinical World Model reframes the field's central question from whether AI works to in which competency coordinates reliability has been demonstrated, and for whom.
We present DeepEarth, a self-supervised multi-modal world model with Earth4D, a novel planetary-scale 4D space-time positional encoder. Earth4D extends 3D multi-resolution hash encoding to include time, efficiently scaling across the planet over centuries with sub-meter, sub-second precision. Multi-modal encoders (e.g. vision-language models) are fused with Earth4D embeddings and trained via masked reconstruction. We demonstrate Earth4D's expressive power by achieving state-of-the-art performance on an ecological forecasting benchmark. Earth4D with learnable hash probing surpasses a multi-modal foundation model pre-trained on substantially more data. Access open source code and download models at: https://github.com/legel/deepearth
While large language models have become the prevailing approach for agentic reasoning and planning, their success in symbolic domains does not readily translate to the physical world. Spatial intelligence, the ability to perceive 3D structure, reason about object relationships, and act under physical constraints, is an orthogonal capability that proves important for embodied agents. Existing surveys address either agentic architectures or spatial domains in isolation. None provide a unified framework connecting these complementary capabilities. This paper bridges that gap. Through a thorough review of over 2,000 papers, citing 742 works from top-tier venues, we introduce a unified three-axis taxonomy connecting agentic capabilities with spatial tasks across scales. Crucially, we distinguish spatial grounding (metric understanding of geometry and physics) from symbolic grounding (associating images with text), arguing that perception alone does not confer agency. Our analysis reveals three key findings mapped to these axes: (1) hierarchical memory systems (Capability axis) are important for long-horizon spatial tasks. (2) GNN-LLM integration (Task axis) is a promising approach for structured spatial reasoning. (3) World models (Scale axis) are essential for safe deployment across micro-to-macro spatial scales. We conclude by identifying six grand challenges and outlining directions for future research, including the need for unified evaluation frameworks to standardize cross-domain assessment. This taxonomy provides a foundation for unifying fragmented research efforts and enabling the next generation of spatially-aware autonomous systems in robotics, autonomous vehicles, and geospatial intelligence.
World models learn to simulate environment dynamics from experience, enabling sample-efficient reinforcement learning. But what do these models actually represent internally? We apply interpretability techniques--including linear and nonlinear probing, causal interventions, and attention analysis--to two architecturally distinct world models: IRIS (discrete token transformer) and DIAMOND (continuous diffusion UNet), trained on Atari Breakout and Pong. Using linear probes, we find that both models develop linearly decodable representations of game state variables (object positions, scores), with MLP probes yielding only marginally higher R^2, confirming that these representations are approximately linear. Causal interventions--shifting hidden states along probe-derived directions--produce correlated changes in model predictions, providing evidence that representations are functionally used rather than merely correlated. Analysis of IRIS attention heads reveals spatial specialization: specific heads attend preferentially to tokens overlapping with game objects. Multi-baseline token ablation experiments consistently identify object-containing tokens as disproportionately important. Our findings provide interpretability evidence that learned world models develop structured, approximately linear internal representations of environment state across two games and two architectures.
This study examines the 2005 reunion of the Presbyterian Church of Korea’s Hapdong and
Gaehyuk denominations as a case of ecclesial reconciliation within modern Korean
Protestantism. While scholarship has often emphasised narratives of schism, this work
reconstructs the historical, theological, and socio-political dynamics that produced repeated
divisions and the pathways that enabled reunification. An integrated theoretical framework
combines historical reconstruction, phenomenological hermeneutics, narrative identity, and
contextualist intellectual history with theological analysis of the church, ecclesial communion,
and reconciliation. Central attention is given to the leadership and repentant agency of Pastor
Kyu-Oh Chung, long regarded as a principal figure behind nearly every major schism within
Korean Presbyterianism. The very leader once associated with division became the agent of
repentance and, by taking a prominent role at the forefront of the unity negotiations, decisively
enabled the success of the 2005 “Union Principles Agreement.” Documentary sources identify
the importance of union committees, mutual recognition among presbyteries and seminaries,
and public reaffirmation of union at the 90th General Assembly.The study illustrates that
reunion was theologically based on biblical themes of reconciliation and unity. It was
institutionally based on negotiated recognition among seminaries, presbyteries, and the
media, and politically in a public reaffirmation that confronted contemporary issues. The work
contends that diversity in confession is compatible with unity. Open government and reciprocal
recognition can sustain doctrinal integrity and cure the wounds inflicted on broken
relationships. It offers an empirically nuanced and theoretically astute examination of Christian
unity in Korea and argues that repentance, leadership, and institutional form function as
essential mediators between theological principle and ecclesial practice. Building on these
findings, it sets forth a mobile approach for post-conflict settlement churches, contending that
humility, theological norms held in common, and incremental organizational integration can
transform difference into durable unity with enduring significance for the global church.
Brennen A. Hill, Mant Koh En Wei, Thangavel Jishnuanandh
Robust coordination is critical for effective decision-making in multi-agent systems, especially under partial observability. A central question in Multi-Agent Reinforcement Learning (MARL) is whether to engineer communication protocols or learn them end-to-end. We investigate this dichotomy using embodied world models. We propose and compare two communication strategies for a cooperative task-allocation problem. The first, Learned Direct Communication (LDC), learns a protocol end-to-end. The second, Intention Communication, uses an engineered inductive bias: a compact, learned world model, the Imagined Trajectory Generation Module (ITGM), which uses the agent's own policy to simulate future states. A Message Generation Network (MGN) then compresses this plan into a message. We evaluate these approaches on goal-directed interaction in a grid world, a canonical abstraction for embodied AI problems, while scaling environmental complexity. Our experiments reveal that while emergent communication is viable in simple settings, the engineered, world model-based approach shows superior performance, sample efficiency, and scalability as complexity increases. These findings advocate for integrating structured, predictive models into MARL agents to enable active, goal-driven coordination.
Remote Sensing Image Dehazing (RSID) poses significant challenges in real-world scenarios due to the complex atmospheric conditions and severe color distortions that degrade image quality. The scarcity of real-world remote sensing hazy image pairs has compelled existing methods to rely primarily on synthetic datasets. However, these methods struggle with real-world applications due to the inherent domain gap between synthetic and real data. To address this, we introduce Real-World Remote Sensing Hazy Image Dataset (RRSHID), the first large-scale dataset featuring real-world hazy and dehazed image pairs across diverse atmospheric conditions. Based on this, we propose MCAF-Net, a novel framework tailored for real-world RSID. Its effectiveness arises from three innovative components: Multi-branch Feature Integration Block Aggregator (MFIBA), which enables robust feature extraction through cascaded integration blocks and parallel multi-branch processing; Color-Calibrated Self-Supervised Attention Module (CSAM), which mitigates complex color distortions via self-supervised learning and attention-guided refinement; and Multi-Scale Feature Adaptive Fusion Module (MFAFM), which integrates features effectively while preserving local details and global context. Extensive experiments validate that MCAF-Net demonstrates state-of-the-art performance in real-world RSID, while maintaining competitive performance on synthetic datasets. The introduction of RRSHID and MCAF-Net sets new benchmarks for real-world RSID research, advancing practical solutions for this complex task. The code and dataset are publicly available at https://github.com/lwCVer/RRSHID.
Recent advances in general-purpose AI systems with attention-based transformers offer a potential window into how the neocortex and cerebellum, despite their relatively uniform circuit architectures, give rise to diverse functions and, ultimately, to human intelligence. This Perspective provides a cross-domain comparison between the brain and AI that goes beyond the traditional focus on visual processing, adopting the emerging perspecive of world-model-based computation. Here, we identify shared computational mechanisms in the attention-based neocortex and the non-attentional cerebellum: both predict future world events from past inputs and construct internal world models through prediction-error learning. These predictive world models are repurposed for seemingly distinct functions -- understanding in sensory processing and generation in motor processing -- enabling the brain to achieve multi-domain capabilities and human-like adaptive intelligence. Notably, attention-based AI has independently converged on a similar learning paradigm and world-model-based computation. We conclude that these shared mechanisms in both biological and artificial systems constitute a core computational foundation for realizing diverse functions including high-level intelligence, despite their relatively uniform circuit structures. Our theoretical insights bridge neuroscience and AI, advancing our understanding of the computational essence of intelligence.
On-the-fly reasoning often requires adaptation to novel problems under limited data and distribution shift. This work introduces CausalARC: an experimental testbed for AI reasoning in low-data and out-of-distribution regimes, modeled after the Abstraction and Reasoning Corpus (ARC). Each CausalARC reasoning task is sampled from a fully specified causal world model, formally expressed as a structural causal model. Principled data augmentations provide observational, interventional, and counterfactual feedback about the world model in the form of few-shot, in-context learning demonstrations. As a proof-of-concept, we illustrate the use of CausalARC for four language model evaluation settings: (1) abstract reasoning with test-time training, (2) counterfactual reasoning with in-context learning, (3) program synthesis, and (4) causal discovery with logical reasoning. Within- and between-model performance varied heavily across tasks, indicating room for significant improvement in language model reasoning.
The capacity of an embodied agent to understand, predict, and interact with its environment is fundamentally contingent on an internal world model. This paper introduces a novel framework for investigating the formation and adaptation of such world models within a biological substrate: human neural organoids. We present a curriculum of three scalable, closed-loop virtual environments designed to train these biological agents and probe the underlying synaptic mechanisms of learning, such as long-term potentiation (LTP) and long-term depression (LTD). We detail the design of three distinct task environments that demand progressively more sophisticated world models for successful decision-making: (1) a conditional avoidance task for learning static state-action contingencies, (2) a one-dimensional predator-prey scenario for goal-directed interaction, and (3) a replication of the classic Pong game for modeling dynamic, continuous-time systems. For each environment, we formalize the state and action spaces, the sensory encoding and motor decoding mechanisms, and the feedback protocols based on predictable (reward) and unpredictable (punishment) stimulation, which serve to drive model refinement. In a significant methodological advance, we propose a meta-learning approach where a Large Language Model automates the generative design and optimization of experimental protocols, thereby scaling the process of environment and curriculum design. Finally, we outline a multi-modal evaluation strategy that moves beyond task performance to directly measure the physical correlates of the learned world model by quantifying synaptic plasticity at electrophysiological, cellular, and molecular levels. This work bridges the gap between model-based reinforcement learning and computational neuroscience, offering a unique platform for studying embodiment, decision-making, and the physical basis of intelligence.
Lingaraj Temple is widely known as a great religious and sacred centre (Kshetra). A place of pilgrimage (tirtha) where Hindu pilgrims from different parts of the country visit and throng wearily, and often barefooted under fatigue and hardship to get the sight of the Lord, which they believe, obliterates the transgressions of one’s whole life. The article deals with the roles of pilgrim guides in Lingaraj Temple. For the comfort of the pilgrims in the temple, there are pilgrim guides functioning among the temple servants (sevayats). These temple functionaries serving the pilgrims are known as Pilgrim Guides or otherwise locally known as ‘Yatri Panda’ or ‘Elaka Panda’. There are three categories of sacred servants (sevayats), Pujapanda, Brahmana (Mahasuar) and Badu who are the Brahmin sevayats and perform rituals in the temple. They act as pilgrim guides of the temple during non-service days. This study is purely an ethnographic study based on fieldwork.
Daesoon Jinrihoe (Korean: 대순진리회) is one of many new religious organizations that emerged in Korea in the second half of the 19th - first half of the 20th century as an alternative to traditional religious and philosophical teachings (Buddhism, Taoism, Confucianism) and imposed Western Christianity. Some of them did not go beyond the closed communities, while the Daesoon Jinrihoe organization, relying on rich traditions of intellectual and social resistance to the authorities and expansion of the West, was able not only to survive in the modern world, but also to become successful and the fastest growing in the Republic of Korea. The purpose of the study is to analyze the ideological roots of Daesoon Jinrihoe, the fundamental aspects of the religious doctrine of the movement and its current state. The main source for studying the doctrine, principles and goals of Daesoon Jinrihoe is the canonical work “Jeongyeong”, first published in 1929 and representing a record of the deeds of Kang Jeungsan, who became, according to religious dogma, the earthly incarnation of the Supreme God Sangje and in this capacity began to reorganize the Universe. “Jeongyeong” consists of several sections written in hanja (Korean writing based on Chinese characters) and describing the life and miracles of Kang Jeungsan in specific life situations. Another source - “The Constitution of Dao” is a list of rights, duties and rules of conduct for members of the organization, describes its internal structure and management system. In domestic historiography there are no academic works devoted to the topic under study. The author relied on the works of Korean and Western researchers, who to one degree or another covered the history of the creation and activities of Daesoon Jinrihoe. The conclusions of this study are that the popularity of the Daesoon teaching is based on its appeal to the ideas and images of traditional Korean religions that have become part of the national mentality, and the inclusion of the organization in the modern national and world agenda.
Western thinkers have often seen Confucianism as unique in that it does not fit well into standard categories of either religion or philosophy. In terms of how this difficulty is reflected in academia, it seems that Confucianism is simultaneously both a religion and a philosophy, and neither. This paper attempts to begin a discussion of a related issue, which has been relatively underappreciated; namely, how does Confucianism attempt to convince people to be Confucian? Restricting our discussion to the Analects, we can find appeals to a belief which seem almost religious, rational arguments which seem philosophical, and a host of other methodologies that may help to convince readers to follow the Confucian way. The discussion of the Analects is cursory here, and only a few passages will be discussed in a general manner. We will contrast this with a general outline of approaches to convincing people found in two-world theories, which can help to illuminate the uniqueness of the Analects, especially when it is read as a one-world perspective.
Anirudh S Chakravarthy, Meghana Reddy Ganesina, Peiyun Hu
et al.
Addressing Lidar Panoptic Segmentation (LPS ) is crucial for safe deployment of autonomous vehicles. LPS aims to recognize and segment lidar points w.r.t. a pre-defined vocabulary of semantic classes, including thing classes of countable objects (e.g., pedestrians and vehicles) and stuff classes of amorphous regions (e.g., vegetation and road). Importantly, LPS requires segmenting individual thing instances (e.g., every single vehicle). Current LPS methods make an unrealistic assumption that the semantic class vocabulary is fixed in the real open world, but in fact, class ontologies usually evolve over time as robots encounter instances of novel classes that are considered to be unknowns w.r.t. the pre-defined class vocabulary. To address this unrealistic assumption, we study LPS in the Open World (LiPSOW): we train models on a dataset with a pre-defined semantic class vocabulary and study their generalization to a larger dataset where novel instances of thing and stuff classes can appear. This experimental setting leads to interesting conclusions. While prior art train class-specific instance segmentation methods and obtain state-of-the-art results on known classes, methods based on class-agnostic bottom-up grouping perform favorably on classes outside of the initial class vocabulary (i.e., unknown classes). Unfortunately, these methods do not perform on-par with fully data-driven methods on known classes. Our work suggests a middle ground: we perform class-agnostic point clustering and over-segment the input cloud in a hierarchical fashion, followed by binary point segment classification, akin to Region Proposal Network [1]. We obtain the final point cloud segmentation by computing a cut in the weighted hierarchical tree of point segments, independently of semantic classification. Remarkably, this unified approach leads to strong performance on both known and unknown classes.
Shashank Hegde, Satyajeet Das, Gautam Salhotra
et al.
With the increasing availability of open-source robotic data, imitation learning has become a promising approach for both manipulation and locomotion. Diffusion models are now widely used to train large, generalized policies that predict controls or trajectories, leveraging their ability to model multimodal action distributions. However, this generality comes at the cost of larger model sizes and slower inference, an acute limitation for robotic tasks requiring high control frequencies. Moreover, Diffusion Policy (DP), a popular trajectory-generation approach, suffers from a trade-off between performance and action horizon: fewer diffusion queries lead to larger trajectory chunks, which in turn accumulate tracking errors. To overcome these challenges, we introduce WARPD (World model Assisted Reactive Policy Diffusion), a method that generates closed-loop policies (weights for neural policies) directly, instead of open-loop trajectories. By learning behavioral distributions in parameter space rather than trajectory space, WARPD offers two major advantages: (1) extended action horizons with robustness to perturbations, while maintaining high task performance, and (2) significantly reduced inference costs. Empirically, WARPD outperforms DP in long-horizon and perturbed environments, and achieves multitask performance on par with DP while requiring only ~ 1/45th of the inference-time FLOPs per step.
Célestin Coquidé, José Lages, Dima L. Shepelyansky
We extend the opinion formation approach to probe the world influence of economical organizations. Our opinion formation model mimics a battle between currencies within the international trade network. Based on the United Nations Comtrade database, we construct the world trade network for the years of the last decade from 2010 to 2020. We consider different core groups constituted by countries preferring to trade in a specific currency. We will consider principally two core groups, namely, 5 Anglo-Saxon countries which prefer to trade in US dollar and the 11 BRICS+ which prefer to trade in a hypothetical currency, hereafter called BRI, pegged to their economies. We determine the trade currency preference of the other countries via a Monte Carlo process depending on the direct transactions between the countries. The results obtained in the frame of this mathematical model show that starting from year 2014 the majority of the world countries would have preferred to trade in BRI than USD. The Monte Carlo process reaches a steady state with 3 distinct groups: two groups of countries preferring, whatever is the initial distribution of the trade currency preferences, to trade, one in BRI and the other in USD, and a third group of countries swinging as a whole between USD and BRI depending on the initial distribution of the trade currency preferences. We also analyze the battle between USD, EUR and BRI, and present the reduced Google matrix description of the trade relations between the Anglo-Saxon countries and the BRICS+.
One of the contended issues in Pentecostal studies is the exclusion or inclusion of African Independent Churches (AICs) as part of the Pentecostal tradition. This article resuscitates this old debate by looking specifically into the inclusion or exclusion of Zionist AICs in the Pentecostal tradition to make a theological contribution. This will be achieved by briefly discussing the Zionist AICs in the South African context. The various factors that contribute to who or what qualifies to be Pentecostal will be discussed by conceptualising Pentecostal identity. In other words, the research question to be answered is on top of which mountain do we stand to include or exclude Zionist AICs in the Pentecostal tradition? Therefore, this article will discuss the theological criteria used to include the Zionist AICs in Pentecostalism. Similarly, the theological criteria for the exclusion of the Zionist AICs will be discussed in detail. The aim and objective of this article are to attempt to answer the question are Zionist AICs Pentecostal? This will have implications for the study of Pentecostal theology in the South African context. Data was collected by reviewing and analysing literature on the AICs and their relationship with the Pentecostal tradition within a theological framework.
The South African Constitution and the law have ensured noticeable progress in
acknowledging the LGBTQI community’s rights. Consequently, there is now a legal framework that protects LGBTQI people, and any discriminatory behaviour and utterances can be prosecuted by law. The struggle now lies within the religious sector, where limited progress has been made. This paper focuses on the progress made within the Christian religion in terms of creating policies and regulations to protect LGBTQI community members’ safety. We focus on same-sex relationships by arguing that even today, such relationships are not openly
accepted by the Church. Using lived religion theory, we revisit Ntombana et al.’s (2020) findings and argue that queer people are closer to God and more spiritual than the homophobic Christians who attend daily Christian fellowship meetings. As queer people are in the minority and oppressed by the church system, we use Tutu and Boesak’s theology to argue that they are closer to God than homophobic Christians. We highlight that during the COVID-19 lockdown, the Christian community suffered, while the queer community flourished because
their spirituality is not based on the Church’s orthodox tradition but on their relationship with God.
Through an analysis of Islamic jurisprudence, ethics, and theological frameworks, the research identifies key considerations and implications for Islamic broadcasters engaging with social media platforms. The study employed a qualitative approach, drawing upon Islamic texts, scholarly literature, and interviews with Islamic broadcasters. Analysis focused on identifying theological principles relevant to social media engagement and examining strategies used by broadcasters to navigate digital communication landscapes. One significant finding is the tension between the imperative to disseminate Islamic teachings widely and the need to maintain theological integrity in online interactions. The study highlights the nuanced approaches taken by Islamic broadcasters to navigate this tension, emphasizing the importance of upholding Islamic values such as tawhid (monotheism) and adab (etiquette) while leveraging the reach and accessibility of social media. Overall, the findings underscore the necessity for a theological re-evaluation of Islamic broadcasting practices in the digital age. By synthesizing traditional Islamic teachings, which refer to the principles, beliefs, and practices that have been passed down in Islam since the time of the Prophet Muhammad SAW and his companions, with contemporary digital realities, this research advocates for an approach that balances the imperative of outreach with the necessity of maintaining theological integrity. It concludes by emphasizing the importance of aligning digital practices with core Islamic values to ensure ethical and responsible engagement with social media platforms.
The Quran, the cornerstone of Islam, underpins the faith, society, and civilization of the Muslim world. As the final divine revelation bestowed upon Prophet Muhammad (SAW), it was his duty to memorize and convey its teachings to his companions. This responsibility extends to the entire Muslim community, especially the ulamas (Islamic scholars), as the Prophet (SAW) emphasized the importance of learning and teaching the Quran. Through the dedicated efforts of ulamas, the Quran has been preserved from any alterations, fulfilling Allah's promise of protection. This research focuses on the translation methodology of Allama Burqaee, a distinguished Persian translator of the Quran. The significance of this study lies in recognizing Allama Burqaee's outstanding contributions to Islamic scholarship. He authored over a hundred works, including translations of the Quran and Hadith, and conducted extensive research on various religions and conflicting beliefs. His tafseer (exegesis) is particularly valuable for students of Islamic studies due to its comprehensive integration of Sunni and Shia scholarly perspectives and ease of access to primary sources. The choice of this topic is motivated by several reasons: (1) To acknowledge Allama Burqaee's efforts in opposing religious innovations (bida'). (2) To understand Allama Burqaee's approach to the Quran, Sunnah, and other Islamic sciences. (3) To highlight his inclusion of diverse scholarly opinions and contemporary ijtihad (intellectual striving). (4) To elucidate his emphasis on the importance of aqeeda (belief), a foundational aspect of Islam. (5) To discuss his call for Muslim unity based on the Quran and Sunnah. The research methodology involves following Allama Burqaee's approach without alterations, providing explanations and critical commentary when necessary. The steps include examining Quranic verses with references, analyzing Hadith and athar (traditions) from primary sources, considering scholarly consensus, and offering concise definitions and examples. This study aims to present a thorough understanding of Allama Burqaee's translation methods and their implications for Islamic scholarship.