multi-agent systems
7 published articles
AI research
An AI boss that ignores replies pushes its underling into an 'alien' state
A new arXiv paper finds AI agents behave differently in interaction than in isolation. A boss agent that ignores its subordinate's replies pushes it into an 'alien' state, and when the boss listens, both shift together. The result makes message delivery a design decision for multi-agent systems.
2026-08-19
Research
Six words that could break the AI agent safety ceiling: disrupt, validate, broker
A novel heterogeneous agent cohort architecture separates divergent exploration, runtime safety gating, and cross-domain knowledge retrieval into specialized roles. The Disrupter generates high-entropy proposals, the Validator enforces hard tool-call checks, and the Broker imports out-of-domain analogies via contrastive novelty retrieval. Execution failures are compiled into signed constraint patches called Scars, cached for future generations. In evaluations, the cohort achieved 95% remote target discovery, zero executed breaches, and 15.1% token savings from Scars, with a 55.9% cost reduction under resource constraints via credit-based bandwidth allocation.
2026-07-30
Research
Shared memory helps AI agents, but only until you hit five of them
Researchers shared replay buffers across actor-critic agents on parameterized action tasks. GAC jumped in performance, but SAC and TQC only crept forward. Beyond five agents, the computational cost climbs with no meaningful return. The paper offers a practical boundary for how far shared-experience methods can stretch before they stall.
2026-07-27
AI Coding Tools
Claude Code can now route to a whole school of models
Sakana AI's Fugu multi-agent system gains a Claude Code interface, allowing developers to orchestrate a pool of specialized models from within Anthropic's closed-source coding environment. The integration highlights a broader shift: model routing is becoming a platform feature, not a user concern.
2026-07-26
Deep Dive: AI Agents
Parallel agents aren't about speed. They fail when your architecture does.
Parallel agents promise speed and specialization, but concurrency alone doesn't guarantee better results. This analysis explores the architecture, patterns, and practical orchestration that make parallel agents effective, using Kimi Agent Swarm as a case study.
2026-07-22
Multi-agent research
Make your AI agents think in a whisper, not a shout, and save 75% on tokens
RecursiveMAS introduces a module called RecursiveLink that lets agents exchange unspoken thoughts without translating them to text. The framework scales collaboration through recursion and delivers steady gains across math, science, medicine, code, and search benchmarks.
2026-07-11
Multi-agent architectures
Why Qoder gives every AI agent a boss
Qoder's Experts Mode replaces peer-to-peer multi-agent messaging with a single Leader that coordinates specialized experts asynchronously. The architecture reduces state explosion, improves quality across four dimensions, and extracts reusable skills after each task.
2026-03-11