As large language models (LLMs) are increasingly deployed as autonomous agents, understanding their cooperation and social mechanisms is becoming increasingly important. In particular, how LLMs balance self-interest and collective well-being is a critical challenge for ensuring alignment, robustness, and safe deployment. In this paper, we examine the challenge of costly sanctioning in multi-agent LLM systems, where an agent must decide whether to invest its own resources to incentivize cooperation or penalize defection. To study this, we adapt a public goods game with institutional choice from behavioral economics, allowing us to observe how different LLMs navigate social dilemmas over repeated interactions. Our analysis reveals four distinct behavioral patterns among models: some consistently establish and sustain high levels of cooperation, others fluctuate between engagement and disengagement, some gradually decline in cooperative behavior over time, and others rigidly follow fixed strategies regardless of outcomes. Surprisingly, we find that reasoning LLMs, such as the o1 series, struggle significantly with cooperation, whereas some traditional LLMs consistently achieve high levels of cooperation. These findings suggest that the current approach to improving LLMs, which focuses on enhancing their reasoning capabilities, does not necessarily lead to cooperation, providing valuable insights for deploying LLM agents in environments that require sustained collaboration. Our code is available at https://github.com/davidguzmanp/SanctSim

Estimating high-dimensional covariance matrices is a key task across many fields. This paper explores the theoretical limits of distributed covariance estimation in a feature-split setting, where communication between agents is constrained. Specifically, we study a scenario in which multiple agents each observe different components of i.i.d. samples drawn from a sub-Gaussian random vector. A central server seeks to estimate the complete covariance matrix using a limited number of bits communicated by each agent. We obtain a nearly tight minimax lower bound for covariance matrix estimation under operator norm and Frobenius norm. Our main technical tool is a novel generalization of the strong data processing inequality (SDPI), termed the Conditional Strong Data Processing Inequality (C-SDPI) coefficient, introduced in this work. The C-SDPI coefficient shares key properties such as tensorization with the conventional SDPI. Crucially, it quantifies the average contraction in a state-dependent channel and can be significantly lower than the worst-case SDPI coefficient over the state input. Utilizing the doubling trick of Geng-Nair and an operator Jensen inequality, we compute this coefficient for Gaussian mixture channels. We then employ it to establish minimax lower bounds on estimation error, capturing the trade-offs among sample size, communication cost, and data dimensionality. Building on this, we present a nearly optimal estimation protocol whose sample and communication requirements match the lower bounds up to logarithmic factors. Unlike much of the existing literature, our framework does not assume infinite samples or Gaussian distributions, making it broadly applicable. Finally, we extend our analysis to interactive protocols, showing interaction can significantly reduce communication requirements compared to non-interactive schemes.

Central to agentic capability and trustworthiness of language model agents (LMAs) is the extent they maintain stable, reliable, identity over time. However, LMAs inherit pathologies from large language models (LLMs) (statelessness, stochasticity, sensitivity to prompts and linguistically-intermediation) which can undermine their identifiability, continuity, persistence and consistency. This attrition of identity can erode their reliability, trustworthiness and utility by interfering with their agentic capabilities such as reasoning, planning and action. To address these challenges, we introduce \textit{agent identity evals} (AIE), a rigorous, statistically-driven, empirical framework for measuring the degree to which an LMA system exhibit and maintain their agentic identity over time, including their capabilities, properties and ability to recover from state perturbations. AIE comprises a set of novel metrics which can integrate with other measures of performance, capability and agentic robustness to assist in the design of optimal LMA infrastructure and scaffolding such as memory and tools. We set out formal definitions and methods that can be applied at each stage of the LMA life-cycle, and worked examples of how to apply them.

Multi-Agent Path Finding (MAPF) is the problem of finding a set of collision-free paths, one for each agent in a shared environment. Its objective is to minimize the sum of path costs (SOC), where the path cost of each agent is defined as the travel time from its start location to its target location. Explicit Estimation Conflict-Based Search (EECBS) is the leading algorithm for bounded-suboptimal MAPF, with the SOC of the solution being at most a user-specified factor $w$ away from optimal. EECBS maintains sets of paths and a lower bound $LB$ on the optimal SOC. Then, it iteratively selects a set of paths whose SOC is at most $w \cdot LB$ and introduces constraints to resolve collisions. For each path in a set, EECBS maintains a lower bound on its optimal path that satisfies constraints. By finding an individually bounded-suboptimal path with cost at most a threshold of $w$ times its lower bound, EECBS guarantees to find a bounded-suboptimal solution. To speed up EECBS, previous work uses flex distribution to increase the threshold. Though EECBS with flex distribution guarantees to find a bounded-suboptimal solution, increasing the thresholds may push the SOC beyond $w \cdot LB$, forcing EECBS to switch among different sets of paths instead of resolving collisions on a particular set of paths, and thus reducing efficiency. To address this issue, we propose Conflict-Based Flex Distribution that distributes flex in proportion to the number of collisions. We also estimate the delays needed to satisfy constraints and propose Delay-Based Flex Distribution. On top of that, we propose Mixed-Strategy Flex Distribution, combining both in a hierarchical framework. We prove that EECBS with our new flex distribution mechanisms is complete and bounded-suboptimal. Our experiments show that our approaches outperform the original (greedy) flex distribution.

Recent advancements in Large Language Models (LLMs) have catalyzed a paradigm shift from static prediction systems to agentic AI agents capable of reasoning, interacting with tools, and adapting to complex tasks. While LLM-based agentic systems have shown promise across many domains, their application to medical imaging remains in its infancy. In this work, we introduce AURA, the first visual linguistic explainability agent designed specifically for comprehensive analysis, explanation, and evaluation of medical images. By enabling dynamic interactions, contextual explanations, and hypothesis testing, AURA represents a significant advancement toward more transparent, adaptable, and clinically aligned AI systems. We highlight the promise of agentic AI in transforming medical image analysis from static predictions to interactive decision support. Leveraging Qwen-32B, an LLM-based architecture, AURA integrates a modular toolbox comprising: (i) a segmentation suite with phase grounding, pathology segmentation, and anatomy segmentation to localize clinically meaningful regions; (ii) a counterfactual image-generation module that supports reasoning through image-level explanations; and (iii) a set of evaluation tools including pixel-wise difference-map analysis, classification, and advanced state-of-the-art components to assess diagnostic relevance and visual interpretability.

Navigation in dynamic environments requires autonomous systems to reason about uncertainties in the behavior of other agents. In this paper, we introduce a unified framework that combines trajectory planning with multimodal predictions and active probing to enhance decision-making under uncertainty. We develop a novel risk metric that seamlessly integrates multimodal prediction uncertainties through mixture models. When these uncertainties follow a Gaussian mixture distribution, we prove that our risk metric admits a closed-form solution, and is always finite, thus ensuring analytical tractability. To reduce prediction ambiguity, we incorporate an active probing mechanism that strategically selects actions to improve its estimates of behavioral parameters of other agents, while simultaneously handling multimodal uncertainties. We extensively evaluate our framework in autonomous navigation scenarios using the MetaDrive simulation environment. Results demonstrate that our active probing approach successfully navigates complex traffic scenarios with uncertain predictions. Additionally, our framework shows robust performance across diverse traffic agent behavior models, indicating its broad applicability to real-world autonomous navigation challenges. Code and videos are available at https://darshangm.github.io/papers/active-probing-multimodal-predictions/.

Large language models (LLMs) turn writing into a live exchange between humans and software. We characterize this new medium as a discursive network that treats people and LLMs as equal nodes and tracks how their statements circulate. We define the generation of erroneous information as invalidation (any factual, logical, or structural breach) and show it follows four hazards: drift from truth, self-repair, fresh fabrication, and external detection. We develop a general mathematical model of discursive networks that shows that a network governed only by drift and self-repair stabilizes at a modest error rate. Giving each false claim even a small chance of peer review shifts the system to a truth-dominant state. We operationalize peer review with the open-source Flaws-of-Others (FOO) algorithm: a configurable loop in which any set of agents critique one another while a harmonizer merges their verdicts. We identify an ethical transgression, epithesis, that occurs when humans fail to engage in the discursive network. The takeaway is practical and cultural: reliability in this new medium comes not from perfecting single models but from connecting imperfect ones into networks that enforce mutual accountability.

Large Language Model (LLM) based multi-agent systems have shown remarkable performance in various tasks, especially when enhanced through collaborative communication. However, current methods often rely on a fixed number of agents and static communication structures, limiting their ability to adapt to varying task complexities. In this paper, we propose Adaptive Graph Pruning (AGP), a novel task-adaptive multi-agent collaboration framework that jointly optimizes agent quantity (hard-pruning) and communication topology (soft-pruning). Specifically, our method employs a two-stage training strategy: firstly, independently training soft-pruning networks for different agent quantities to determine optimal agent-quantity-specific complete graphs and positional masks across specific tasks; and then jointly optimizing hard-pruning and soft-pruning within a maximum complete graph to dynamically configure the number of agents and their communication topologies per task. Extensive experiments demonstrate that our approach is: (1) High-performing, achieving state-of-the-art results across six benchmarks and consistently generalizes across multiple mainstream LLM architectures, with a increase in performance of $2.58\%\sim 9.84\%$; (2) Task-adaptive, dynamically constructing optimized communication topologies tailored to specific tasks, with an extremely high performance in all three task categories (general reasoning, mathematical reasoning, and code generation); (3) Token-economical, having fewer training steps and token consumption at the same time, with a decrease in token consumption of $90\%+$; and (4) Training-efficient, achieving high performance with very few training steps compared with other methods. The performance will surpass the existing baselines after about ten steps of training under six benchmarks.

Affordances - i.e. possibilities for action that an environment or objects in it provide - are important for robots operating in human environments to perceive. Existing approaches train such capabilities on annotated static images or shapes. This work presents a novel dataset for affordance learning of common household tasks. Unlike previous approaches, our dataset consists of video sequences demonstrating the tasks from first- and third-person perspectives, along with metadata about the affordances that are manifested in the task, and is aimed towards training perception systems to recognize affordance manifestations. The demonstrations were collected from several participants and in total record about seven hours of human activity. The variety of task performances also allows studying preparatory maneuvers that people may perform for a task, such as how they arrange their task space, which is also relevant for collaborative service robots.

--This work tackles the physical layer security (PLS) problem of maximizing the secrecy rate in heterogeneous UA V networks (HetUA VNs) under propulsion energy constraints. Unlike prior studies that assume uniform UA V capabilities or overlook energy-security trade-offs, we consider a realistic scenario where UA Vs with diverse payloads and computation resources collaborate to serve ground terminals in the presence of eavesdroppers. T o manage the complex coupling between UA V motion and communication, we propose a hierarchical optimization framework. The inner layer uses a semidefinite relaxation (SDR)-based S2DC algorithm combining penalty functions and difference-of-convex (d.c.) programming to solve the secrecy precoding problem with fixed UA V positions. The outer layer introduces a Large Language Model (LLM)-guided heuristic multi-agent reinforcement learning approach (LLM-HeMARL) for trajectory optimization. LLM-HeMARL efficiently incorporates expert heuristics policy generated by the LLM, enabling UA Vs to learn energy-aware, security-driven trajectories without the inference overhead of real-time LLM calls. The simulation results show that our method outperforms existing baselines in secrecy rate and energy efficiency, with consistent robustness across varying UA V swarm sizes and random seeds. ITH the rapid advancement of 6G technology, unmanned aerial vehicles (UA Vs) have increasingly become a critical component of modern communication infrastructure, owing to their high mobility, strong scalability, and the provision of reliable line-of-sight (LoS) links [1], [2]. However, the broadcast nature of wireless channels over LoS links makes UA V communications more susceptible to eavesdropping and jamming attacks compared to traditional terrestrial networks, which poses significant security and privacy threats.