multi-agent system
PulseAugur coverage of multi-agent system — every cluster mentioning multi-agent system across labs, papers, and developer communities, ranked by signal.
18 day(s) with sentiment data
-
AI prompt caching failure fixed by reordering message context
A developer discovered that their multi-agent AI system was not benefiting from prompt caching due to the order of messages in their API calls. Prompt caching systems typically match on a prefix of the input, and by pla…
-
New frameworks aim to standardize evaluation of multi-agent AI collaboration
Two new research papers introduce frameworks for evaluating multi-agent systems (MAS) built on large language models (LLMs). The first, ForestBench, proposes a unified graph framework to map heterogeneous execution trac…
-
New framework tackles hidden Byzantine attacks in multi-agent systems
Researchers have developed a new theoretical framework and algorithm for online security learning in cooperative multi-agent systems facing hidden Byzantine attacks. The study identifies that an attacker's information a…
-
New framework for agent-native organizations separates persistent and dynamic layers
This paper introduces a framework for designing agent-native organizations that separates operational layers into persistent and dynamic components. The persistent layer includes a four-store record architecture and spe…
-
New model links agent mobility, memory, and network structure to convention tipping points
A new agent-based model explores how mobility, memory, and network structure influence convention tipping points. Researchers found that in many scenarios, a committed minority can inevitably drive a population to adopt…
-
HELENA framework enhances multi-agent systems with novel coordination
Researchers have introduced HELENA, a novel multi-agent system framework designed to enhance analytical capacity by integrating diverse reasoning paths while mitigating noise. HELENA constructs a composite graph from co…
-
Researchers adapt multi-agent attacks/defenses for LLM recommender systems
Researchers have adapted attacks and defenses from general multi-agent systems (MAS) to collaborative filtering (CF) systems powered by autonomous language agents. The study evaluated these adaptations under varying con…
-
New research explores LLM agents trained with symbolic options and cooperative evolution strategies
Researchers are exploring new methods for training large language models (LLMs) to act as agents in complex environments. One approach, detailed in a new arXiv paper, uses a "frontier coding model" to generate symbolic …
-
New backdoor vulnerability discovered in LLM multi-agent systems
Researchers have identified a new vulnerability in LLM-based multi-agent systems where collaboration can inadvertently trigger backdoor behavior. This occurs when collective evidence from multiple agents reaches a hidde…
-
New framework enhances AI agent verifiability in digital twins
Researchers have introduced a novel neuro-symbolic decentralized governance framework designed to enhance the verifiability of AI agents within digital twin ecosystems. This framework integrates probabilistic neural rea…
-
Intellibooks offers framework for data-driven enterprise AI strategy
Intellibooks has introduced a practical framework designed to help enterprises build a data-driven strategy, even when implementing advanced AI technologies like Generative AI or RAG. The framework, detailed in their 'D…
-
New MARS framework enables autonomous repair for multi-agent systems
Researchers have developed MARS, a novel search-based framework for autonomously repairing multi-agent systems (MAS). MARS formulates the repair process as a Monte Carlo Tree Search (MCTS) problem, navigating potential …
-
New research paper introduces procedural fairness for multi-agent bandit systems
A new paper introduces the concept of procedural fairness to multi-agent multi-armed bandit systems, moving beyond outcome-based fairness metrics. The research proposes formalizing procedural fairness as equal voice and…
-
Adversarial comments bypass LLM code vulnerability detectors
Researchers have developed ALIBI, a novel attack framework that inserts adversarial natural-language comments into source code to bypass LLM-based vulnerability detectors. This technique successfully manipulates over 90…
-
AI visualization of oral histories faces challenges in narrative preservation
A new research paper explores the challenges of using generative AI to visualize oral history interviews, particularly for diaspora communities. The study proposes a framework with 15 metrics to evaluate AI-generated vi…
-
SafeFlow framework enhances multi-agent system security against malicious propagation
Researchers have introduced SafeFlow, a new framework designed to enhance security in multi-agent systems. This system addresses the challenge of malicious intent being fragmented across specialized agents, which can le…
-
Paper explores emergent behavior in financial markets using formal methods
A new paper explores emergent behavior in financial markets, drawing parallels between complex systems and the formal methods community. The research identifies and structures sources of complexity within electronic fin…
-
AI Agent Memory: An Architectural Challenge, Not Just a Model Problem
A new paper argues that the memory systems of AI agents are fundamentally an architectural problem, not simply a matter of model capabilities or context window size. The author proposes a lifecycle approach to memory, b…
-
LLM framework enhances accuracy in identifying adverse drug events
A new research paper details a human-in-the-loop framework utilizing a retrieval-augmented, multi-agent large language model (LLM) to identify cutaneous immune-related adverse events (cirAEs) from clinical notes. This L…
-
Network perturbations can distort agreement trajectories in multi-agent systems
A new paper published on arXiv details how network perturbations can significantly alter agreement trajectories in Linear Time-Invariant (LTI) multi-agent systems. The research introduces a Laplace-domain criterion to a…