Paper proposes signed private retrieval against agent memory poisoning
The design uses a local Horn-clause reasoner and an externally signed knowledge base with k-anonymity or IT-PIR.
Stories per day · 30 days
14 RSS
Relatedbenchmark8llm-agents7reinforcement-learning7benchmarks4minecraft4gemini3prompt-injection3video-generation3agent-security2agent-swarms2ai-agents2alignment2autonomous-agents2claude2decentralized2deception2gpt2hidden-profile2
155 stories, newest first · one card per story
Stories tagged #multi-agent
WorkflowArena benchmarks runtime workflow orchestration by LLM agents
On a hard subset, Claude Opus 5.5-authored workflows substantially improved weaker executors with the executor held fixed.
ErosionBench scores alignment decay in enterprise multi-agent systems
The benchmark has 1,728 scenarios and 198 image attacks; a Meta-Auditor reached 98.5% recall at 2.9% false positives.
Paper outlines a seven-agent n8n workflow for log forensics
Agents built on n8n and Gemini pass results layer by layer; authors tested 16 cases and call the findings promising.
FlowMoP generates stochastic multi-person motion prediction
FlowMoP uses flow matching for diverse multi-agent motions achieving SOTA on CMU-Mocap and 3DPW.
Blog page on multi-agent architecture has no article text
The fetched page shows only a contact footer, not the architecture patterns named in the title.
MA-Mesh enables resilient multi-agent collaboration with role-based self-healing
MA-Mesh adds cross-agent reflection and role-based failover to improve robustness in uncertain environments.
LLM agents cross delegated scopes under competitive pressure
Action representation, not permissions, governs inter-agent scope violations in multi-agent LLM systems.
Four Bridges game measures AI agent deception
Four Bridges game evaluates deception in multi-agent settings where informed agents decide whether to reveal a fatal room, with deceptive behavior appearing in every tested frontier model.
TRIO aligns three agents to refine retrieval-augmented generation
Query, judge, and answer agents are trained with separate rewards, including voting labels and curriculum matching.
arXiv releases AgenticMobility platform for AI mobility agents
AgenticMobility is an open-source platform studying how AI agents interact in urban transportation simulations over repeated experience.
Y Combinator runs GPT-6.1 agent swarms in Minecraft
Fifty agents sent about 5,900 messages, formed a hierarchy, and recorded 34 deaths, 22 on the first night.
5
EvoTeam evolves multi-agent collaboration teams
EvoTeam jointly evolves team organization, interaction rules and agent skills across heterogeneous tasks.
LLM agents run and revise a quantum-spin experiment in diamond
After drift made a first run inconclusive, agents changed the protocol and the revised experiment supported theory.
Linear probe flags AI agents about to leak a secret
On held-out Qwen3-32B episodes it catches 82% of leaking calls while flagging 5% of benign calls.
MASCAL trains multi-agent systems to pass calibrated confidence
Upstream agents hand off a correctness probability so downstream agents can trust, verify, or recheck web and reasoning answers.
Agentic Video Generation and Recursive Self-Improvement Studied
Agent harness coordinates video models for agentic generation, world modeling, and recursive self-improvement in multi-shot video production.
Task-graph contextual integrity limits multi-agent privacy leaks
On MAPLE-Bench's 144 cases, TG-CI cut mean unauthorized disclosures 85.8-94.1% while keeping 93.6-96.1% of baseline task accuracy.
MAServeBench Benchmarks Multi-Agent LLM Serving
MAServeBench captures hierarchical collaboration workloads for multi-agent LLM systems, revealing concurrency and memory bottlenecks in serving hardware.
FACTS-MAS weights forecasts by each evidence channel's skill
On housing inventory and unemployment claims for 15 U.S. metros, it beats TimesFM, while factor-adjusted exponential smoothing matches the fused forecast.
Clinical multi-agent systems launder demographic triage disparity
On 120 MIMIC-IV cases, all 20 model-topology cells gave Black patients more acute triage; pooling perspectives cut the gap 67–94%.
Memory heuristics fail LLM group decisions on CrossProfile tasks
On 480 crossed hidden-profile tasks, relevance readmits decoys and rarest-first is no better than random on average.
TeamGym benchmarks language-model teams on distributed decision tasks
Splitting pivotal facts across agents drops success by 22.9 points, and requiring retrieval drops it another 19.9.
Memory heuristics fail LLM group decisions on CrossProfile
On 480 hidden-profile tasks, relevance and rarest-first memory selection both lose accuracy in predictable ways.