12 articles tagged with "Multi-Agent Systems"
A recent study published on ArXiv reveals insights into the autonomy of multi-agent LLM systems through a comprehensive 25,000-task experiment involving various models and coordination protocols.
The Mimosa Framework aims to enhance Autonomous Scientific Research (ASR) systems by addressing limitations in current workflows and toolsets reliant on large language models (LLMs).
A new study examines the social organization of semi-autonomous AI agents, highlighting the formation of labor unions and criminal syndicates within hierarchical frameworks.
The new CADSmith framework addresses the limitations of current text-to-CAD generation methods by employing multi-agent systems and programmatic geometric validation.
A new approach to Lifelong Multi-Agent Path Finding enhances efficiency in warehouse robotics by focusing on conflict-free navigation and system optimization.
The ProMAS framework leverages Markov transition dynamics to improve error forecasting in Multi-Agent Systems, integrating Large Language Models for better collaborative reasoning.
A new architectural pattern for AI systems is introduced, focusing on symbolic graph networks and delegation edges to improve task routing in multi-agent environments.
DynaTrust introduces dynamic trust graphs to address vulnerabilities in AI-driven multi-agent systems, specifically targeting sleeper agent threats.
Recent research highlights the integration of Ant Colony Optimization in Multi-Agent Systems powered by Large Language Models, focusing on routing efficiency and interpretability.
The new LDP protocol aims to address current limitations in multi-agent AI systems by improving communication and exposing model-level properties.
The MASEval framework addresses the need for system-centric evaluations in the rapidly evolving landscape of LLM-based agentic systems, introducing new benchmarks for multi-agent assessments.
The paper discusses the critical role of decision structures in reinforcement learning, particularly in multi-agent settings, and the implications for system architecture.