Multi-agent systems often rely on specialized agents with distinct roles rather than general-purpose agents that perform the entire task independently. However, the conditions that govern the optimal degree of specialization remain poorly understood. In this work, we propose that specialist teams outperform generalist ones when environmental constraints limit task parallelizability -- the potential to execute task components concurrently. Drawing inspiration from distributed systems, we introduce a heuristic to predict the relative efficiency of generalist versus specialist teams by estimating the speed-up achieved when two agents perform a task in parallel rather than focus on complementary subtasks. We validate this heuristic through three multi-agent reinforcement learning (MARL) experiments in Overcooked-AI, demonstrating that key factors limiting task parallelizability influence specialization. We also observe that as the state space expands, agents tend to converge on specialist strategies, even when generalist ones are theoretically more efficient, highlighting potential biases in MARL training algorithms. Our findings provide a principled framework for interpreting specialization given the task and environment, and introduce a novel benchmark for evaluating whether MARL finds optimal strategies.

The promising potential of AI and network convergence in improving networking performance and enabling new service capabilities has recently attracted significant interest. Existing network AI solutions, while powerful, are mainly built based on the close-loop and passive learning framework, resulting in major limitations in autonomous solution finding and dynamic environmental adaptation. Agentic AI has recently been introduced as a promising solution to address the above limitations and pave the way for true generally intelligent and beneficial AI systems. The key idea is to create a networking ecosystem to support a diverse range of autonomous and embodied AI agents in fulfilling their goals. In this paper, we focus on the novel challenges and requirements of agentic AI networking. We propose AgentNet, a novel framework for supporting interaction, collaborative learning, and knowledge transfer among AI agents. We introduce a general architectural framework of AgentNet and then propose a generative foundation model (GFM)-based implementation in which multiple GFM-as-agents have been created as an interactive knowledge-base to bootstrap the development of embodied AI agents according to different task requirements and environmental features. We consider two application scenarios, digital-twin-based industrial automation and metaverse-based infotainment system, to describe how to apply AgentNet for supporting efficient task-driven collaboration and interaction among AI agents.

This paper presents a novel hierarchical, safety-critical control framework that integrates distributed nonlinear model predictive controllers (DNMPCs) with control barrier functions (CBFs) to enable cooperative locomotion of multi-agent quadrupedal robots in complex environments. While NMPC-based methods are widely adopted for enforcing safety constraints and navigating multi-robot systems (MRSs) through intricate environments, ensuring the safety of MRSs requires a formal definition grounded in the concept of invariant sets. CBFs, typically implemented via quadratic programs (QPs) at the planning layer, provide formal safety guarantees. However, their zero-control horizon limits their effectiveness for extended trajectory planning in inherently unstable, underactuated, and nonlinear legged robot models. Furthermore, the integration of CBFs into real-time NMPC for sophisticated MRSs, such as quadrupedal robot teams, remains underexplored. This paper develops computationally efficient, distributed NMPC algorithms that incorporate CBF-based collision safety guarantees within a consensus protocol, enabling longer planning horizons for safe cooperative locomotion under disturbances and rough terrain conditions. The optimal trajectories generated by the DNMPCs are tracked using full-order, nonlinear whole-body controllers at the low level. The proposed approach is validated through extensive numerical simulations with up to four Unitree A1 robots and hardware experiments involving two A1 robots subjected to external pushes, rough terrain, and uncertain obstacle information. Comparative analysis demonstrates that the proposed CBF-based DNMPCs achieve a 27.89% higher success rate than conventional NMPCs without CBF constraints.

Escape rooms present a unique cognitive challenge that demands exploration-driven planning: players should actively search their environment, continuously update their knowledge based on new discoveries, and connect disparate clues to determine which elements are relevant to their objectives. Motivated by this, we introduce VisEscape, a benchmark of 20 virtual escape rooms specifically designed to evaluate AI models under these challenging conditions, where success depends not only on solving isolated puzzles but also on iteratively constructing and refining spatial-temporal knowledge of a dynamically changing environment. On VisEscape, we observed that even state-of-the-art multimodal models generally fail to escape the rooms, showing considerable variation in their levels of progress and trajectories. To address this issue, we propose VisEscaper, which effectively integrates Memory, Feedback, and ReAct modules, demonstrating significant improvements by performing 3.7 times more effectively and 5.0 times more efficiently on average.

Co-speech gestures convey a wide variety of meanings and play an important role in face-to-face human interactions. These gestures significantly influence the addressee's engagement, recall, comprehension, and attitudes toward the speaker. Similarly, they impact interactions between humans and embodied virtual agents. The process of selecting and animating meaningful gestures has thus become a key focus in the design of these agents. However, automating this gesture selection process poses a significant challenge. Prior gesture generation techniques have varied from fully automated, data-driven methods, which often struggle to produce contextually meaningful gestures, to more manual approaches that require crafting specific gesture expertise and are time-consuming and lack generalizability. In this paper, we leverage the semantic capabilities of Large Language Models to develop a gesture selection approach that suggests meaningful, appropriate co-speech gestures. We first describe how information on gestures is encoded into GPT-4. Then, we conduct a study to evaluate alternative prompting approaches for their ability to select meaningful, contextually relevant gestures and to align them appropriately with the co-speech utterance. Finally, we detail and demonstrate how this approach has been implemented within a virtual agent system, automating the selection and subsequent animation of the selected gestures for enhanced human-agent interactions.

Recent LLM-based agent frameworks have demonstrated impressive capabilities in task delegation and workflow orchestration, but face significant challenges in maintaining context awareness and ensuring planning consistency. This paper presents SagaLLM, a structured multi-agent framework that addresses four fundamental limitations in current LLM approaches: inadequate self-validation, context narrowing, lacking transaction properties, and insufficient inter-agent coordination. By implementing specialized context management agents and validation protocols, SagaLLM preserves critical constraints and state information throughout complex planning processes, enabling robust and consistent decision-making even during disruptions. We evaluate our approach using selected problems from the REALM benchmark, focusing on sequential and reactive planning scenarios that challenge both context retention and adaptive reasoning. Our experiments with state-of-the-art LLMs, Claude 3.7, DeepSeek R1, GPT-4o, and GPT-o1, demonstrate that while these models exhibit impressive reasoning capabilities, they struggle with maintaining global constraint awareness during complex planning tasks, particularly when adapting to unexpected changes. In contrast, the distributed cognitive architecture of SagaLLM shows significant improvements in planning consistency, constraint enforcement, and adaptation to disruptions in various scenarios.

This paper introduces a framework for human swarm interaction studies that measures situation awareness in dynamic environments. A tablet-based interface was developed for a user study by implementing the concepts introduced in the framework, where operators guided a robotic swarm in a single-target search task, marking hazardous cells unknown to the swarm. Both subjective and objective situation awareness measures were used, with task performance evaluated based on how close the robots were to the target. The framework enabled a structured investigation of the role of situation awareness in human swarm interaction, leading to key findings such as improved task performance across attempts, showing the interface was learnable, centroid active robot position proved to be a useful task performance metric for assessing situation awareness, perception and projection played a key role in task performance, highlighting their importance in interface design and both subjective and objective situation awareness influenced task performance, emphasizing the need for interfaces that support both. These findings validate our framework as a structured approach for integrating situation awareness concepts into human swarm interaction studies, offering a systematic way to assess situation awareness and task performance. The framework can be applied to other swarming studies to evaluate interface learnability, identify meaningful task performance metrics, and refine interface designs to enhance situation awareness, ultimately improving human swarm interaction in dynamic environments.

Graphical user interface (GUI) has become integral to modern society, making it crucial to be understood for human-centric systems. However, unlike natural images or documents, GUIs comprise artificially designed graphical elements arranged to convey specific semantic meanings. Current multi-modal large language models (MLLMs) already proficient in processing graphical and textual components suffer from hurdles in GUI understanding due to the lack of explicit spatial structure modeling. Moreover, obtaining high-quality spatial structure data is challenging due to privacy issues and noisy environments. To address these challenges, we present MP-GUI, a specially designed MLLM for GUI understanding. MP-GUI features three precisely specialized perceivers to extract graphical, textual, and spatial modalities from the screen as GUI-tailored visual clues, with spatial structure refinement strategy and adaptively combined via a fusion gate to meet the specific preferences of different GUI understanding tasks. To cope with the scarcity of training data, we also introduce a pipeline for automatically data collecting. Extensive experiments demonstrate that MP-GUI achieves impressive results on various GUI understanding tasks with limited data.

Despite rapid progress on AI benchmarks, the real-world meaning of benchmark performance remains unclear. To quantify the capabilities of AI systems in terms of human capabilities, we propose a new metric: 50%-task-completion time horizon. This is the time humans typically take to complete tasks that AI models can complete with 50% success rate. We first timed humans with relevant domain expertise on a combination of RE-Bench, HCAST, and 66 novel shorter tasks. On these tasks, current frontier AI models such as Claude 3.7 Sonnet have a 50% time horizon of around 50 minutes. Furthermore, frontier AI time horizon has been doubling approximately every seven months since 2019, though the trend may have accelerated in 2024. The increase in AI models' time horizons seems to be primarily driven by greater reliability and ability to adapt to mistakes, combined with better logical reasoning and tool use capabilities. We discuss the limitations of our results -- including their degree of external validity -- and the implications of increased autonomy for dangerous capabilities. If these results generalize to real-world software tasks, extrapolation of this trend predicts that within 5 years, AI systems will be capable of automating many software tasks that currently take humans a month.

While human cognition inherently retrieves information from diverse and specialized knowledge sources during decision-making processes, current Retrieval-Augmented Generation (RAG) systems typically operate through single-source knowledge retrieval, leading to a cognitive-algorithmic discrepancy. To bridge this gap, we introduce MoK-RAG, a novel multi-source RAG framework that implements a mixture of knowledge paths enhanced retrieval mechanism through functional partitioning of a large language model (LLM) corpus into distinct sections, enabling retrieval from multiple specialized knowledge paths. Applied to the generation of 3D simulated environments, our proposed MoK-RAG3D enhances this paradigm by partitioning 3D assets into distinct sections and organizing them based on a hierarchical knowledge tree structure. Different from previous methods that only use manual evaluation, we pioneered the introduction of automated evaluation methods for 3D scenes. Both automatic and human evaluations in our experiments demonstrate that MoK-RAG3D can assist Embodied AI agents in generating diverse scenes.