In this paper, we investigate a reconfigurable intelligent surface (RIS)-aided integrated sensing and communication (ISAC) system, where the RIS is modeled using multiport network theory based on theZ-parameter representation. Unlike conventional RIS models based on reflection-coefficient matrices, we characterize RIS reconfigurability with tunable circuit impedances, thus capturing the electromagnetic mutual coupling (MC) effects among RIS elements. Specifically, we jointly optimize the transmit covariance matrix at the base station (BS) and the RIS tunable load impedance matrix to maximize the radar signal-to-noise ratio (SNR). The power budget at the BS and the quality of service (QoS) constraints for the communication users are also satisfied. To highlight the impact of electromagnetic MC on the system, both the no-MC and the MC-aware cases are considered. For the non-convex MC-aware problem, we propose an alternating optimization (AO) algorithm that integrates Neumann series approximation, semidefinite relaxation (SDR), and sequential rank-one constraint relaxation (SROCR) techniques. As a simplified form of the MC-aware case, the no-MC case can be considered as a sub-algorithm embedded in the proposed solution framework. Numerical results show that electromagnetic MC significantly affects the system performance, especially under sub-wavelength spacing.

The powerful capabilities of Large Language Models (LLMs) have led to their growing use in evaluating human-generated content, particularly in evaluating research ideas within academic settings. Existing solutions primarily rely on prompt-based LLM methods or fine-tuned lightweight language models for idea evaluation. However, these methods are often unstable and struggle to comprehend the complex semantic information embedded in the ideas, impeding their ability to perform high-quality evaluations. To address the above challenges, we propose $\texttt{GraphEval}$, a lightweight graph-based LLM framework for idea evaluation. Our insight is that a complex idea can be broken down into comprehensible viewpoint nodes using prompts from small LLMs. These viewpoint nodes can then be linked together through edges created from LLM-based relation extraction and/or BERT similarity scores. The created viewpoint-graph can be used to conveniently propagate scores across view-nodes to improve the robustness of the idea evaluations. In particular, we propose two lightweight graph-based methods for idea evaluation: (1) GraphEval-LP: a training-free label propagation algorithm that propagates evaluation scores from known view-nodes to unknown nodes; (2) GraphEval-GNN: a Graph Neural Networks (GNN) that is trained to predict the evaluation scores given the observed graph with minimal computation resources. Moreover, to overcome LLM's limitation in objectively assessing the novelty of ideas, we further propose a novelty detection model to GraphEval-GNN to enhance its capability in judging idea novelty. Experiments on two datasets show $\texttt{GraphEval}$ improves F1 scores by at least 14% with low computation and API costs. Additionally, $\texttt{GraphEval}$ can effectively detect plagiarized ideas.

Building agents with adaptive behavior in cooperative tasks stands as a paramount goal in the realm of multi-agent systems. Current approaches to developing cooperative agents rely primarily on learning-based methods, whose policy generalization depends heavily on the diversity of teammates they interact with during the training phase. Such reliance, however, constrains the agents' capacity for strategic adaptation when cooperating with unfamiliar teammates, which becomes a significant challenge in zero-shot coordination scenarios. To address this challenge, we propose ProAgent, a novel framework that harnesses large language models (LLMs) to create proactive agents capable of dynamically adapting their behavior to enhance cooperation with teammates. ProAgent can analyze the present state, and infer the intentions of teammates from observations. It then updates its beliefs in alignment with the teammates' subsequent actual behaviors. Moreover, ProAgent exhibits a high degree of modularity and interpretability, making it easily integrated into various of coordination scenarios. Experimental evaluations conducted within the Overcooked-AI environment unveil the remarkable performance superiority of ProAgent, outperforming five methods based on self-play and population-based training when cooperating with AI agents. Furthermore, in partnered with human proxy models, its performance exhibits an average improvement exceeding 10% compared to the current state-of-the-art method. For more information about our project, please visit https://pku-proagent.github.io.

The performance of mean square error (MSE) in channel estimation for multi-cell massive multiple-input multiple-output (MIMO) systems with Rician fading is studied. In this report, we initially derive the closed-form expressions of the probability distribution function and cumulative distribution function of MSE, which are applicable for any number of base-station antennas M and any Rician K-factor. Furthermore, we perform an asymptotic analysis for both strong line-of-sight (LOS) and Rayleigh fading scenarios. Subsequently, we present closed-form expressions for the expectation of MSE $\left(\operatorname{Exp}_{\mathrm{mse}}\right)$ and the variance of MSE. Next, utilizing maximal-ratio combining detector, we investigate the relationship between achievable downlink spectral efficiency and $\operatorname{Exp}_{\mathrm{mse}}$. It is observed that as $\operatorname{Exp}_{\text {mse }}$ increases, the achievable downlink spectral efficiency constantly reduces, eventually reaches a given constant. Finally, Monte-Carlo simulations are performed to corroborate the results discussed earlier.