This introduces a fundamental challenge in establishing computational trust, specifically how one agent can verify that another's output was genuinely produced by a claimed LLM, and not falsified or generated by a cheaper or inferior model. T o address this challenge, this paper proposes a verification framework that achieves tractable asymmetric effort, where the cost to verify a computation is substantially lower than the cost to perform it. Our approach is built upon the principle of deterministic replicability, a property inherent to autoregressive models that strictly necessitates a computationally homogeneous environment where all agents operate on identical hardware and software stacks. Within this defined context, our framework enables multiple validators to probabilistically audit small, random segments of an LLM's output and it distributes the verification workload effectively. The simulations demonstrated that targeted verification can be over 12 times faster than full regeneration, with tunable parameters to adjust the detection probability. By establishing a tractable mechanism for auditable LLM systems, our work offers a foundational layer for responsible AI and serves as a cornerstone for future research into the more complex, heterogeneous multi-agent systems.

Multi-agent debate (MAD) is an emerging approach to improving the reasoning capabilities of large language models (LLMs). Existing MAD methods rely on multiple rounds of interaction among agents to reach consensus, and the final output is selected by majority voting in the last round. However, this consensus-based design faces several limitations. First, multiple rounds of communication increases token overhead and limits scalability. Second, due to the inherent conformity of LLMs, agents that initially produce correct responses may be influenced by incorrect ones during the debate process, causing error propagation. Third, majority voting introduces randomness and unfairness in the decision-making phase, and can degrade the reasoning performance. To address these issues, we propose \textsc{Free-MAD}, a novel MAD framework that eliminates the need for consensus among agents. \textsc{Free-MAD} introduces a novel score-based decision mechanism that evaluates the entire debate trajectory rather than relying on the last round only. This mechanism tracks how each agent's reasoning evolves, enabling more accurate and fair outcomes. In addition, \textsc{Free-MAD} reconstructs the debate phase by introducing anti-conformity, a mechanism that enables agents to mitigate excessive influence from the majority. Experiments on eight benchmark datasets demonstrate that \textsc{Free-MAD} significantly improves reasoning performance while requiring only a single-round debate and thus reducing token costs. We also show that compared to existing MAD approaches, \textsc{Free-MAD} exhibits improved robustness in real-world attack scenarios.

Agent-based models (ABMs) are gaining increasing traction in several domains, due to their ability to represent complex systems that are not easily expressible with classical mathematical models. This expressivity and richness come at a cost: ABMs can typically be analyzed only through simulation, making their analysis challenging. Specifically, when studying the output of ABMs, the analyst is often confronted with practical questions such as: (i) how many independent replications should be run? (ii) how many initial time steps should be discarded as a warm-up? (iii) after the warm-up, how long should the model run? (iv) what are the right parameter values? Analysts usually resort to rules of thumb and experimentation, which lack statistical rigor. This is mainly because addressing these points takes time, and analysts prefer to spend their limited time improving the model. In this paper, we propose a methodology, drawing on the field of Statistical Model Checking, to automate the process and provide guarantees of statistical rigor for ABMs written in NetLogo, one of the most popular ABM platforms. We discuss MultiVeStA, a tool that dramatically reduces the time and human intervention needed to run statistically rigorous checks on ABM outputs, and introduce its integration with NetLogo. Using two ABMs from the NetLogo library, we showcase MultiVeStA's analysis capabilities for NetLogo ABMs, as well as a novel application to statistically rigorous calibration. Our tool-chain makes it immediate to perform statistical checks with NetLogo models, promoting more rigorous and reliable analyses of ABM outputs.

Abstract--This paper presents an agent-based simulation model for coordinating battery recharging in drone swarms, focusing on applications in Internet of Things (IoT) and Industry 4.0 environments. The proposed model includes a detailed description of the simulation methodology, system architecture, and implementation. One practical use case is explored: Smart Farming, highlighting how autonomous coordination strategies can optimize battery usage and mission efficiency in large-scale drone deployments. This work uses a machine learning technique to analyze the agent-based simulation sensitivity analysis output results. Drones have become important tools within the Internet of Things, and can be used in agribusiness, disaster response, logistics, and other usages.

The dark patterns, deceptive interface designs manipulating user behaviors, have been extensively studied for their effects on human decision-making and autonomy. Yet, with the rising prominence of LLM-powered GUI agents that automate tasks from high-level intents, understanding how dark patterns affect agents is increasingly important. We present a two-phase empirical study examining how agents, human participants, and human-AI teams respond to 16 types of dark patterns across diverse scenarios. Phase 1 highlights that agents often fail to recognize dark patterns, and even when aware, prioritize task completion over protective action. Phase 2 revealed divergent failure modes: humans succumb due to cognitive shortcuts and habitual compliance, while agents falter from procedural blind spots. Human oversight improved avoidance but introduced costs such as attentional tunneling and cognitive load. Our findings show neither humans nor agents are uniformly resilient, and collaboration introduces new vulnerabilities, suggesting design needs for transparency, adjustable autonomy, and oversight.

Human language is one of the most expressive tools for conveying intent, yet most artificial or biological systems lack mechanisms to interpret or respond meaningfully to it. Bridging this gap could enable more natural forms of control over complex, decentralized systems. In AI and artificial life, recent work explores how language can specify high-level goals, but most systems still depend on engineered rewards, task-specific supervision, or rigid command sets, limiting generalization to novel instructions. Similar constraints apply in synthetic biology and bioengineering, where the locus of control is often genomic rather than environmental perturbation. A key open question is whether artificial or biological collectives can be guided by free-form natural language alone, without task-specific tuning or carefully designed evaluation metrics. We provide one possible answer here by showing, for the first time, that simple agents' collective behavior can be guided by free-form language prompts: one AI model transforms an imperative prompt into an intervention that is applied to simulated cells; a second AI model scores how well the prompt describes the resulting cellular dynamics; and the former AI model is evolved to improve the scores generated by the latter. Unlike previous work, our method does not require engineered fitness functions or domain-specific prompt design. We show that the evolved system generalizes to unseen prompts without retraining. By treating natural language as a control layer, the system suggests a future in which spoken or written prompts could direct computational, robotic, or biological systems to desired behaviors. This work provides a concrete step toward this vision of AI-biology partnerships, in which language replaces mathematical objective functions, fixed rules, and domain-specific programming.

For groups of autonomous agents to achieve a particular goal, they must engage in coordination and long-horizon reasoning. However, designing reward functions to elicit such behavior is challenging. In this paper, we study how self-supervised goal-reaching techniques can be leveraged to enable agents to cooperate. The key idea is that, rather than have agents maximize some scalar reward, agents aim to maximize the likelihood of visiting a certain goal. This problem setting enables human users to specify tasks via a single goal state rather than implementing a complex reward function. While the feedback signal is quite sparse, we will demonstrate that self-supervised goal-reaching techniques enable agents to learn from such feedback. On MARL benchmarks, our proposed method outperforms alternative approaches that have access to the same sparse reward signal as our method. While our method has no explicit mechanism for exploration, we observe that self-supervised multi-agent goal-reaching leads to emergent cooperation and exploration in settings where alternative approaches never witness a single successful trial.

--Wireless communication is evolving into an agent era, where large-scale agents with inherent embodied intelligence are not just users but active participants. The perfect combination of wireless communication and embodied intelligence can achieve a synergetic empowerment and greatly facilitate the development of agent communication. An overview of this synergetic empowerment is presented, framing it as a co-evolutionary process that transforms wireless communication from a simple utility into the digital nervous system of a collective intelligence, while simultaneously elevating isolated agents into a unified superorganism with emergent capabilities far exceeding individual contributions. Furthermore, critical open issues and future research directions are identified. IRELESS communication is evolving into the agent era, marking a fundamental shift from connecting passive information endpoints to enabling massive-scale agent collaboration. Unlike traditional devices, these agents such as autonomous vehicles, industrial robots, and advanced environmental sensors possess inherent embodied intelligence, empowering them to actively perceive, reason, and physically interact with their surroundings [1]. The scale of this transformation is unprecedented. The projections for 2030 estimate that the number of connected IoT devices will reach 125 billion, while monthly global mobile traffic is expected to increase to over 5000 exabytes, representing an 80-fold increase from 2020 [2]. More critically, a growing portion of these devices is the embodied agents that require real-time coordination for complex collective tasks, marking a qualitative shift from isolated sensors to collaborative swarms. Diao and Q. Wu are with the College of Electronic and Information Engineering, Nanjing University of Aeronautics and Astronautics, Nanjing, 210000, P .

Agentic artificial intelligence (AI) -- multi-agent systems that combine large language models with external tools and autonomous planning -- are rapidly transitioning from research laboratories into high-stakes domains. Our earlier "Basic" paper introduced a five-axis framework and proposed preliminary metrics such as goal drift and harm reduction but did not provide an algorithmic instantiation or empirical evidence. This "Advanced" sequel fills that gap. First, we revisit recent benchmarks and industrial deployments to show that technical metrics still dominate evaluations: a systematic review of 84 papers from 2023--2025 found that 83% report capability metrics while only 30% consider human-centred or economic axes [2]. Second, we formalise an Adaptive Multi-Dimensional Monitoring (AMDM) algorithm that normalises heterogeneous metrics, applies per-axis exponentially weighted moving-average thresholds and performs joint anomaly detection via the Mahalanobis distance [7]. Third, we conduct simulations and real-world experiments. AMDM cuts anomaly-detection latency from 12.3 s to 5.6 s on simulated goal drift and reduces false-positive rates from 4.5% to 0.9% compared with static thresholds. We present a comparison table and ROC/PR curves, and we reanalyse case studies to surface missing metrics. Code, data and a reproducibility checklist accompany this paper to facilitate replication. The code supporting this work is available at https://github.com/Manishms18/Adaptive-Multi-Dimensional-Monitoring.

Beyond scratch coding, exploiting large-scale code repositories (e.g., GitHub) for practical tasks is vital in real-world software development, yet current benchmarks rarely evaluate code agents in such authentic, workflow-driven scenarios. To bridge this gap, we introduce GitTaskBench, a benchmark designed to systematically assess this capability via 54 realistic tasks across 7 modalities and 7 domains. Each task pairs a relevant repository with an automated, human-curated evaluation harness specifying practical success criteria. Beyond measuring execution and task success, we also propose the alpha-value metric to quantify the economic benefit of agent performance, which integrates task success rates, token cost, and average developer salaries. Experiments across three state-of-the-art agent frameworks with multiple advanced LLMs show that leveraging code repositories for complex task solving remains challenging: even the best-performing system, OpenHands+Claude 3.7, solves only 48.15% of tasks (recent progress has pushed the frontier further, with RepoMaster+Claude 3.5 achieving a new record of 62.96%). Error analysis attributes over half of failures to seemingly mundane yet critical steps like environment setup and dependency resolution, highlighting the need for more robust workflow management and increased timeout preparedness. By releasing GitTaskBench, we aim to drive progress and attention toward repository-aware code reasoning, execution, and deployment -- moving agents closer to solving complex, end-to-end real-world tasks. The benchmark and code are open-sourced at https://github.com/QuantaAlpha/GitTaskBench.