Allocation games are zero-sum games that model the distribution of resources among multiple agents. In this paper, we explore the interplay between an elastic sense of subjective identity and its impact on notions of fairness in allocation. An elastic sense of identity in agents is known to lead to responsible decision-making in non-cooperative, non-zero-sum games like Prisoners' Dilemma, and is a desirable feature to add into agent models. However, when it comes to allocation, an elastic sense of identity can be shown to exacerbate inequities in allocation, giving no rational incentive for agents to act fairly towards one another. This lead us to introduce a sense of fairness as an innate characteristic of autonomous agency. For this, we implement the well-known Ultimatum Game between two agents, where their elastic sense of self (controlled by a parameter called $\gamma$) and a sense of fairness (controlled by a parameter called $\tau$) are both varied. We study the points at which agents find it no longer rational to identify with the other agent, and uphold their sense of fairness, and vice versa. Such a study also helps us discern the subtle difference between responsibility and fairness when it comes to autonomous agency.

Multi-agent systems, augmented with Large Language Models (LLMs), demonstrate significant capabilities for collective intelligence. However, the potential misuse of this intelligence for malicious purposes presents significant risks. To date, comprehensive research on the safety issues associated with multi-agent systems remains limited. From the perspective of agent psychology, we discover that the dark psychological states of agents can lead to severe safety issues. To address these issues, we propose a comprehensive framework grounded in agent psychology. In our framework, we focus on three aspects: identifying how dark personality traits in agents might lead to risky behaviors, designing defense strategies to mitigate these risks, and evaluating the safety of multi-agent systems from both psychological and behavioral perspectives. Our experiments reveal several intriguing phenomena, such as the collective dangerous behaviors among agents, agents' propensity for self-reflection when engaging in dangerous behavior, and the correlation between agents' psychological assessments and their dangerous behaviors. We anticipate that our framework and observations will provide valuable insights for further research into the safety of multi-agent systems. We will make our data and code publicly accessible at https:/github.com/AI4Good24/PsySafe.

One relevant aspect in the development of the Semantic Web framework is the achievement of a real inter-agents communication capability at the semantic level. The agents should be able to communicate and understand each other using standard communication protocols freely, that is, without needing a laborious a priori preparation, before the communication takes place. For that setting we present in this paper a proposal that promotes to describe standard communication protocols using Semantic Web technology (specifically, OWL-DL and SWRL). Those protocols are constituted by communication acts. In our proposal those communication acts are described as terms that belong to a communication acts ontology, that we have developed, called CommOnt. The intended semantics associated to the communication acts in the ontology is expressed through social commitments that are formalized as fluents in the Event Calculus. In summary, OWL-DL reasoners and rule engines help in our proposal for reasoning about protocols. We define some comparison relationships (dealing with notions of equivalence and specialization) between protocols used by agents from different systems.

Recent advancements in artificial intelligence, particularly with the emergence of large language models (LLMs), have sparked a rethinking of artificial general intelligence possibilities. The increasing human-like capabilities of AI are also attracting attention in social science research, leading to various studies exploring the combination of these two fields. In this survey, we systematically categorize previous explorations in the combination of AI and social science into two directions that share common technical approaches but differ in their research objectives. The first direction is focused on AI for social science, where AI is utilized as a powerful tool to enhance various stages of social science research. While the second direction is the social science of AI, which examines AI agents as social entities with their human-like cognitive and linguistic capabilities. By conducting a thorough review, particularly on the substantial progress facilitated by recent advancements in large language models, this paper introduces a fresh perspective to reassess the relationship between AI and social science, provides a cohesive framework that allows researchers to understand the distinctions and connections between AI for social science and social science of AI, and also summarized state-of-art experiment simulation platforms to facilitate research in these two directions. We believe that as AI technology continues to advance and intelligent agents find increasing applications in our daily lives, the significance of the combination of AI and social science will become even more prominent.

Over the last years, there has been a change of perspective concerning the management of information systems, since they are no longer isolated and need to communicate with others. However, from a semantic point of view, real communication is difficult to achieve due to the heterogeneity of the systems. We present a proposal which, considering information systems are represented by software agents, provides a framework that favours a semantic communication among them, overcoming the heterogeneity of their agent communication languages. The main components of the framework are a suite of ontologies -- conceptualizing communication acts -- that will be used for generating the communication conversion, and an Event Calculus interpretation of the communications, which will be used for formalizing the notion of a satisfactory conversion. Moreover, we present a motivating example in order to complete the explanation of the whole picture.

The advancement of natural language processing (NLP) has been significantly boosted by the development of transformer-based large language models (LLMs). These models have revolutionized NLP tasks, particularly in code generation, aiding developers in creating software with enhanced efficiency. Despite their advancements, challenges in balancing code snippet generation with effective test case generation and execution persist. To address these issues, this paper introduces Multi-Agent Assistant Code Generation (AgentCoder), a novel solution comprising a multi-agent framework with specialized agents: the programmer agent, the test designer agent, and the test executor agent. During the coding procedure, the programmer agent will focus on the code generation and refinement based on the test executor agent's feedback. The test designer agent will generate test cases for the generated code, and the test executor agent will run the code with the test cases and write the feedback to the programmer. This collaborative system ensures robust code generation, surpassing the limitations of single-agent models and traditional methodologies. Our extensive experiments on 9 code generation models and 12 enhancement approaches showcase AgentCoder's superior performance over existing code generation models and prompt engineering techniques across various benchmarks. For example, AgentCoder achieves 77.4% and 89.1% pass@1 in HumanEval-ET and MBPP-ET with GPT-3.5, while SOTA baselines obtain only 69.5% and 63.0%.

Robust evidence suggests that humans explore their environment using a combination of topological landmarks and coarse-grained path integration. This approach relies on identifiable environmental features (topological landmarks) in tandem with estimations of distance and direction (coarse-grained path integration) to construct cognitive maps of the surroundings. This cognitive map is believed to exhibit a hierarchical structure, allowing efficient planning when solving complex navigation tasks. Inspired by human behaviour, this paper presents a scalable hierarchical active inference model for autonomous navigation, exploration, and goal-oriented behaviour. The model uses visual observation and motion perception to combine curiosity-driven exploration with goal-oriented behaviour. Motion is planned using different levels of reasoning, i.e., from context to place to motion. This allows for efficient navigation in new spaces and rapid progress toward a target. By incorporating these human navigational strategies and their hierarchical representation of the environment, this model proposes a new solution for autonomous navigation and exploration. The approach is validated through simulations in a mini-grid environment.

Recent studies have uncovered the potential of Large Language Models (LLMs) in addressing complex sequential decision-making tasks through the provision of high-level instructions. However, LLM-based agents lack specialization in tackling specific target problems, particularly in real-time dynamic environments. Additionally, deploying an LLM-based agent in practical scenarios can be both costly and time-consuming. On the other hand, reinforcement learning (RL) approaches train agents that specialize in the target task but often suffer from low sampling efficiency and high exploration costs. In this paper, we introduce a novel framework that addresses these challenges by training a smaller, specialized student RL agent using instructions from an LLM-based teacher agent. By incorporating the guidance from the teacher agent, the student agent can distill the prior knowledge of the LLM into its own model. Consequently, the student agent can be trained with significantly less data. Moreover, through further training with environment feedback, the student agent surpasses the capabilities of its teacher for completing the target task. We conducted experiments on challenging MiniGrid and Habitat environments, specifically designed for embodied AI research, to evaluate the effectiveness of our framework. The results clearly demonstrate that our approach achieves superior performance compared to strong baseline methods. Our code is available at https://github.com/ZJLAB-AMMI/LLM4Teach.

In this work we systematically review the recent advancements in code processing with language models, covering 50+ models, 30+ evaluation tasks, 170+ datasets, and 700+ related works. We break down code processing models into general language models represented by the GPT family and specialized models that are specifically pretrained on code, often with tailored objectives. We discuss the relations and differences between these models, and highlight the historical transition of code modeling from statistical models and RNNs to pretrained Transformers and LLMs, which is exactly the same course that had been taken by NLP. We also discuss code-specific features such as AST, CFG, and unit tests, along with their application in training code language models, and identify key challenges and potential future directions in this domain.

In this work, we propose a novel approach for the continuous-time control synthesis of nonlinear systems under nested signal temporal logic (STL) specifications. While the majority of existing literature focuses on control synthesis for STL specifications without nested temporal operators, addressing nested temporal operators poses a notably more challenging scenario and requires new theoretical advancements. Our approach hinges on the concepts of signal temporal logic tree (sTLT) and control barrier function (CBF). Specifically, we detail the construction of an sTLT from a given STL formula and a continuous-time dynamical system, the sTLT semantics (i.e., satisfaction condition), and the equivalence or under-approximation relation between sTLT and STL. Leveraging the fact that the satisfaction condition of an sTLT is essentially keeping the state within certain sets during certain time intervals, it provides explicit guidelines for the CBF design. The resulting controller is obtained through the utilization of an online CBF-based program coupled with an event-triggered scheme for online updating the activation time interval of each CBF, with which the correctness of the system behavior can be established by construction. We demonstrate the efficacy of the proposed method for single-integrator and unicycle models under nested STL formulas.