I. de ZarzÃ

I. de ZarzÃ, J. de Curtò, Jordi Cabot et al.

Large Language Models (LLMs) increasingly serve as autonomous reasoning agents in decision support, scientific problem-solving, and multi-agent coordination systems. However, deploying LLM agents in consequential applications requires assurance that their reasoning remains stable under semantically equivalent input variations, a property we term semantic invariance. Standard benchmark evaluations, which assess accuracy on fixed, canonical problem formulations, fail to capture this critical reliability dimension. To address this shortcoming, in this paper we present a metamorphic testing framework for systematically assessing the robustness of LLM reasoning agents, applying eight semantic-preserving transformations (identity, paraphrase, fact reordering, expansion, contraction, academic context, business context, and contrastive formulation) across seven foundation models spanning four distinct architectural families: Hermes (70B, 405B), Qwen3 (30B-A3B, 235B-A22B), DeepSeek-R1, and gpt-oss (20B, 120B). Our evaluation encompasses 19 multi-step reasoning problems across eight scientific domains. The results reveal that model scale does not predict robustness: the smaller Qwen3-30B-A3B achieves the highest stability (79.6% invariant responses, semantic similarity 0.91), while larger models exhibit greater fragility.

J. de Curtò, I. de ZarzÃ

Large Language Models (LLMs) can generate persuasive influence strategies that shift cooperative behavior in multi-agent populations, but a critical question remains: does the resulting cooperation reflect genuine prosocial alignment, or does it mask erosion of agent autonomy, epistemic integrity, and distributional fairness? We introduce Constitutional Multi-Agent Governance (CMAG), a two-stage framework that interposes between an LLM policy compiler and a networked agent population, combining hard constraint filtering with soft penalized-utility optimization that balances cooperation potential against manipulation risk and autonomy pressure. We propose the Ethical Cooperation Score (ECS), a multiplicative composite of cooperation, autonomy, integrity, and fairness that penalizes cooperation achieved through manipulative means. In experiments on scale-free networks of 80 agents under adversarial conditions (70% violating candidates), we benchmark three regimes: full CMAG, naive filtering, and unconstrained optimization. While unconstrained optimization achieves the highest raw cooperation (0.873), it yields the lowest ECS (0.645) due to severe autonomy erosion (0.867) and fairness degradation (0.888). CMAG attains an ECS of 0.741, a 14.9% improvement, while preserving autonomy at 0.985 and integrity at 0.995, with only modest cooperation reduction to 0.770. The naive ablation (ECS = 0.733) confirms that hard constraints alone are insufficient. Pareto analysis shows CMAG dominates the cooperation-autonomy trade-off space, and governance reduces hub-periphery exposure disparities by over 60%. These findings establish that cooperation is not inherently desirable without governance: constitutional constraints are necessary to ensure that LLM-mediated influence produces ethically stable outcomes rather than manipulative equilibria.

J. de Curtò, Adrianne Schneider, Ricardo Yanez et al.

The design of distributed autonomous systems for operation beyond reliable ground contact presents a fundamental tension: as round-trip communication latency grows, the set of decisions delegable to ground operators shrinks. This paper establishes a unified computational methodology for quantifying and comparing this constraint across seven heterogeneous mission architectures, spanning Earth low-orbit surveillance constellations, Mars orbital navigation systems, autonomous underwater mine-clearing swarms, deep-space inter-satellite link networks, and outer-planet in-situ buoy platforms. We introduce the Autonomy Necessity Score, a log-domain latency metric mapping each system continuously from the ground-dependent to the fully-autonomous regime, grounded in nine independently validated computational studies covering Walker spherical-cap coverage mechanics, infrared Neyman-Pearson detection, Extended Kalman Filter hypersonic tracking, cross-mission RF and acoustic link budgets spanning seven orders of magnitude in range, Monte Carlo science-yield sensitivity for TDMA inter-satellite protocols, cross-architecture power budget sizing, distributed magnetic-signature formation emulation, and Arrhenius-corrected cryogenic swarm reliability. Building on this foundation, we evaluate an LLM-based Autonomous Mission Decision Support layer in which three foundation models (Llama-3.3-70B, DeepSeek-V3, and Qwen3-A22B) are queried live via the Nebius AI Studio API across ten structured anomaly scenarios derived directly from the preceding analyses. The best-performing model achieves 80% decision accuracy against physics-grounded ground truth, with all 180 inference calls completing within a 2 s latency budget consistent with radiation-hardened edge deployment, establishing the viability of foundation models as an onboard cognitive layer for high-ANS missions.