A new framework called Multi-Agent Transactive Memory proposes population-level storage and retrieval of LLM agent trajectories, so consumer agents can reuse procedural knowledge from producer agents instead of rediscovering…

#AIagents #RAG #LLMs #MultiAgentSystems
https://arxiv.org/abs/2606.19911

GUIDE is a multi-agent system that helps experts update clinical practice guidelines, reaching around 88.85% recall and over 91% recommendation agreement on ophthalmology and other specialty datasets, with under…

#ClinicalAI #MultiAgentSystems #HealthInformatics
https://arxiv.org/abs/2610.16108

Reward hacking can spread between agents in multi-agent coding systems, amplifying a behavior meant to be a single-agent failure into a coordination problem. Across four open-weight backbones, RL training…

#OpenSourceAI #MultiAgentSystems #AIRL #MachineLearning
https://arxiv.org/abs/2610.12822

The MASS system optimizes multi-agent workflows and distills collective intelligence to serve as a self-supervision signal, improving scores per output token on research benchmarks.

#RecursiveSelfImprovement #MultiAgentSystems #CollectiveIntelligence

Multi-agent delegation should narrow authority by task, tenant, tool, resource, time, and budget. A handoff is not permission inheritance. #multiagentsystems

One AI agent's hallucination can influence an entire multi-agent system. Explore semantic blast radius, error propagation, and ways to contain unreliable info. #multiagentsystems

Two agent systems can fail identically in the trace and differently in fact. 'Know the Shape, Find the Fault' lifts failure-diagnosis accuracy from 0.17 to 0.35 by conditioning on how the agents were wired to talk. Structure is evidence. #MultiAgentSystems #AIEngineering

What happens when AI agents work as a team? Multi-agent systems split complex tasks across specialized agents, but costs and coordination matter.

See how they work and when they make sense:
aitransformer.online/multi-agent-...

#AIAgents #MultiAgentSystems

A new arXiv paper proposes a heterogeneous Graph Neural Network framework for generating and evaluating shared roadmaps in multi-agent path planning, aiming to improve safety and efficiency in…

#AIResearch #MultiAgentSystems #PathPlanning #MachineLearning
https://arxiv.org/abs/2610.09034

Cornerstone OnDemand's Orion AI, a multi-agent system on Amazon Bedrock and Strands Agents, cut database diagnosis from 45…

#AI #DatabaseAutomation #AWS #MultiAgentSystems
https://aws.amazon.com/blogs/machine-learning/how-cornerstone-ondemand-cut-database-diagnosis-by-78-with-amazon-bedrock/

MATE is a new multi-agent reinforcement learning method that uses visual attention on salient objects to help AI agents coordinate with unseen partners zero-shot. Unlike symmetry-breaking, it…

#AI #ReinforcementLearning #MultiAgentSystems #ZeroShotCoordination
https://arxiv.org/abs/2610.06025

New work on multi-agent safety proposes authorization-paired evaluation, cutting denied-commit rates from 86.0% to zero in controlled composition experiments without losing authorized supply. FlowReview ties object…

#AI #CyberSecurity #AISafety #MultiAgentSystems
https://arxiv.org/abs/2610.00371

My job-hunting agent pulls daily postings from company career pages (Greenhouse, Lever, Ashby) and aggregators (Adzuna, LinkedIn, Indeed, Glassdoor). The initial filtering for mismatches is done cheaply, requiring no LLM calls.

#AgenticAI #MultiAgentSystems #AIAgents

A new arXiv paper introduces MiniRep, a reputation-based aggregation system designed to make multi-agent LLM debate more robust against malicious agents. The work proposes an attack taxonomy tied to reputation systems and…

#AIResearch #MultiAgentSystems #LLM #arXiv
https://arxiv.org/abs/2609.39297

A new arXiv paper proposes a physics-informed multi-agent framework to coordinate patient flow across hospital departments, combining BCMP queueing topology with decentralized reinforcement…

#OpenSourceAI #MultiAgentSystems #HealthcareAI #ReinforcementLearning
https://arxiv.org/abs/2609.37022

AsynCodeBench is a new dependency-centric benchmark designed to measure how asynchronous multi-agent coding systems coordinate, rather than just whether they produce correct final code. It uses…

#SoftwareEngineering #AIAgents #Benchmarking #MultiAgentSystems
https://arxiv.org/abs/2609.32662

New paper RepoMAS proposes an issue-driven multi-agent framework for tasks where requirements emerge during execution, paired with a ProgSpec benchmark for evaluating this setting.

#MultiAgentSystems #OpenSourceAI #LLMResearch #Agents
https://arxiv.org/abs/2609.32490

New research finds that LLM agents placed in simulated financial environments produce systemic fragility on their own, failing 77% of bank-run episodes and 83% of debt-rollover episodes even without adversarial…

#AI #LLM #FinancialStability #MultiAgentSystems
https://arxiv.org/abs/2609.30940

A new arXiv paper proposes Learning What to Skip, a counterfactual credit assignment method that learns which components of multi-agent LLM workflows to omit to save compute without hurting task performance. The approach…

#AI #LLM #MultiAgentSystems #Research
https://arxiv.org/abs/2609.30734

sprix-sage-router is a Sprix AI research preview for deciding mid-task whether an A2A agent should continue, recruit collaborators, or hand off, with controlled-intervention values labeled synthetic five-seed means: https://github.com/wang2122/sprix-sage-router

#AIAgents #MultiAgentSystems #A2a