As AI increasingly shapes daily life, energy consumption and data privacy have become pressing concerns. However, the expanding scale of modern neural networks creates a major obstacle for on-device training. Although prior work has concentrated on compact convolutional architectures, we instead apply subspace-based training to transformer models. Motivated by the idea that a model's essential information lies in a fixed subspace, we introduce Weight-Activation Subspace Iteration (W ASI), a method that mitigates the memory bottleneck of backpropagation and boosts inference efficiency in transformer models by restricting training to this subspace. Our results demonstrate that W ASI maintains accuracy comparable to vanilla training while reducing memory usage by up to 62 and computational cost (FLOPs) by up to 2 . On a Raspberry Pi 5, W ASI achieves roughly 1.5 faster training and inference than vanilla training. On-device learning has recently emerged as a promising research direction, enabling deep learning models to be fine-tuned directly on resource-constrained edge devices. This approach addresses critical issues such as privacy and energy consumption, improves scalability, and places control of AI capabilities directly "in user's hands" (Dhar et al., 2021). Prior work on on-device learning has largely focused on vision tasks using convolutional neural network models, primarily because of their compact architectures (Lin et al., 2022; Nguyen et al., 2024; Y ang et al., 2023b; Qu elennec et al., 2024; Bragagnolo et al., 2022; Nguyen et al., 2025). In many real-world applications, however, transformer-based models have become the de facto choice due to their unique architectural mechanisms (V aswani et al., 2017).

Recent surges in LLM-driven intelligent systems largely overlook decades of foundational multi-agent systems (MAS) research, resulting in frameworks with critical limitations such as centralization and inadequate trust and communication protocols. This paper introduces the Fetch.ai architecture, an industrial-strength platform designed to bridge this gap by facilitating the integration of classical MAS principles with modern AI capabilities. We present a novel, multi-layered solution built on a decentralized foundation of on-chain blockchain services for verifiable identity, discovery, and transactions. This is complemented by a comprehensive development framework for creating secure, interoperable agents, a cloud-based platform for deployment, and an intelligent orchestration layer where an agent-native LLM translates high-level human goals into complex, multi-agent workflows. We demonstrate the deployed nature of this system through a decentralized logistics use case where autonomous agents dynamically discover, negotiate, and transact with one another securely. Ultimately, the Fetch.ai stack provides a principled architecture for moving beyond current agent implementations towards open, collaborative, and economically sustainable multi-agent ecosystems.

To succeed in common digital tasks such as web navigation, agents must carry out a variety of specialized tasks such as searching for products or planning a travel route. To tackle these tasks, agents can bootstrap themselves by learning task-specific skills online through interaction with the web environment. In this work, we demonstrate that programs are an effective representation for skills. We propose agent skill induction (ASI), which allows agents to adapt themselves by inducing, verifying, and utilizing program-based skills on the fly. We start with an evaluation on the WebArena agent benchmark and show that ASI outperforms the static baseline agent and its text-skill counterpart by 23.5% and 11.3% in success rate, mainly thanks to the programmatic verification guarantee during the induction phase. ASI also improves efficiency by reducing 10.7-15.3% of the steps over baselines, by composing primitive actions (e.g., click) into higher-level skills (e.g., search product). We then highlight the efficacy of ASI in remaining efficient and accurate under scaled-up web activities. Finally, we examine the generalizability of induced skills when transferring between websites, and find that ASI can effectively reuse common skills, while also updating incompatible skills to versatile website changes.

On-device learning has emerged as a promising direction for AI development, particularly because of its potential to reduce latency issues and mitigate privacy risks associated with device-server communication, while improving energy efficiency. Despite these advantages, significant memory and computational constraints still represent major challenges for its deployment. Drawing on previous studies on low-rank decomposition methods that address activation memory bottlenecks in backpropagation, we propose a novel shortcut approach as an alternative. Our analysis and experiments demonstrate that our method can reduce activation memory usage, even up to $120.09\times$ compared to vanilla training, while also reducing overall training FLOPs up to $1.86\times$ when evaluated on traditional benchmarks.

As Artificial Intelligence (AI) advances toward Artificial General Intelligence (AGI) and eventually Artificial Superintelligence (ASI), it may potentially surpass human control, deviate from human values, and even lead to irreversible catastrophic consequences in extreme cases. This looming risk underscores the critical importance of the "superalignment" problem - ensuring that AI systems which are much smarter than humans, remain aligned with human (compatible) intentions and values. While current scalable oversight and weak-to-strong generalization methods demonstrate certain applicability, they exhibit fundamental flaws in addressing the superalignment paradigm - notably, the unidirectional imposition of human values cannot accommodate superintelligence's autonomy or ensure AGI/ASI's stable learning. We contend that the values for sustainable symbiotic society should be co-shaped by humans and living AI together, achieving "Super Co-alignment." Guided by this vision, we propose a concrete framework that integrates external oversight and intrinsic proactive alignment. External oversight superalignment should be grounded in human-centered ultimate decision, supplemented by interpretable automated evaluation and correction, to achieve continuous alignment with humanity's evolving values. Intrinsic proactive superalignment is rooted in a profound understanding of the Self, others, and society, integrating self-awareness, self-reflection, and empathy to spontaneously infer human intentions, distinguishing good from evil and proactively prioritizing human well-being. The integration of externally-driven oversight with intrinsically-driven proactive alignment will co-shape symbiotic values and rules through iterative human-ASI co-alignment, paving the way for achieving safe and beneficial AGI and ASI for good, for human, and for a symbiotic ecology.

The recent leap in AI capabilities, driven by big generative models, has sparked the possibility of achieving Artificial General Intelligence (AGI) and further triggered discussions on Artificial Superintelligence (ASI), a system surpassing all humans across all domains. This gives rise to the critical research question of: If we realize ASI, how do we align it with human values, ensuring it benefits rather than harms human society, a.k.a., the Superalignment problem. Despite ASI being regarded by many as solely a hypothetical concept, in this paper, we argue that superalignment is achievable and research on it should advance immediately, through simultaneous and alternating optimization of task competence and value conformity. We posit that superalignment is not merely a safeguard for ASI but also necessary for its realization. To support this position, we first provide a formal definition of superalignment rooted in the gap between capability and capacity and elaborate on our argument. Then we review existing paradigms, explore their interconnections and limitations, and illustrate a potential path to superalignment centered on two fundamental principles. We hope this work sheds light on a practical approach for developing the value-aligned next-generation AI, garnering greater benefits and reducing potential harms for humanity.

As Artificial General Intelligence edges closer to reality, Artificial Superintelligence does too. This paper argues that ASI's unparalleled capabilities might lead people to attribute godlike infallibility to it, resulting in a cognitive bias toward unquestioning acceptance of its decisions. By drawing parallels between ASI and divine attributes such as omnipotence, omniscience, and omnipresence, this analysis highlights the risks of conflating technological advancement with moral and ethical superiority. Such dynamics could engender a technocratic theocracy, where decision-making is abdicated to ASI, undermining human agency and critical thinking.

Existing approaches to Theory of Mind (ToM) in Artificial Intelligence (AI) overemphasize prompted, or cue-based, ToM, which may limit our collective ability to develop Artificial Social Intelligence (ASI). Drawing from research in computer science, cognitive science, and related disciplines, we contrast prompted ToM with what we call spontaneous ToM -- reasoning about others' mental states that is grounded in unintentional, possibly uncontrollable cognitive functions. We argue for a principled approach to studying and developing AI ToM and suggest that a robust, or general, ASI will respond to prompts \textit{and} spontaneously engage in social reasoning.

The conventional discourse on existential risks (x-risks) from AI typically focuses on abrupt, dire events caused by advanced AI systems, particularly those that might achieve or surpass human-level intelligence. These events have severe consequences that either lead to human extinction or irreversibly cripple human civilization to a point beyond recovery. This discourse, however, often neglects the serious possibility of AI x-risks manifesting incrementally through a series of smaller yet interconnected disruptions, gradually crossing critical thresholds over time. This paper contrasts the conventional "decisive AI x-risk hypothesis" with an "accumulative AI x-risk hypothesis." While the former envisions an overt AI takeover pathway, characterized by scenarios like uncontrollable superintelligence, the latter suggests a different causal pathway to existential catastrophes. This involves a gradual accumulation of critical AI-induced threats such as severe vulnerabilities and systemic erosion of econopolitical structures. The accumulative hypothesis suggests a boiling frog scenario where incremental AI risks slowly converge, undermining resilience until a triggering event results in irreversible collapse. Through systems analysis, this paper examines the distinct assumptions differentiating these two hypotheses. It is then argued that the accumulative view reconciles seemingly incompatible perspectives on AI risks. The implications of differentiating between these causal pathways -- the decisive and the accumulative -- for the governance of AI risks as well as long-term AI safety are discussed.

In the context of deep learning research, where model introductions continually occur, the need for effective and efficient evaluation remains paramount. Existing methods often emphasize accuracy metrics, overlooking stability. To address this, the paper introduces the Accuracy-Stability Index (ASI), a quantitative measure incorporating both accuracy and stability for assessing deep learning models. Experimental results demonstrate the application of ASI, and a 3D surface model is presented for visualizing ASI, mean accuracy, and coefficient of variation. This paper addresses the important issue of quantitative benchmarking metrics for deep learning models, providing a new approach for accurately evaluating accuracy and stability of deep learning models. The paper concludes with discussions on potential weaknesses and outlines future research directions.