Beyond Individual Intelligence: Surveying Collaboration, Failure Attribution, and Self-Evolution in LLM-based Multi-Agent Systems
:::info Stub — Full Engineering Breakdown Coming This paper has a linked code implementation and was featured on Hugging Face Papers with 44 upvotes. A full breakdown with production viability rating, implementation notes, and honest limitations is being written. Subscribe to AI Letters → :::
| Authors | Shihao Qi et al. |
| Year | 2026 |
| HF Upvotes | 44 |
| arXiv | 2605.14892 |
| Download | |
| Code | https://github.com/mira-ai-lab/awesome-mas-life |
Abstract
LLM-based autonomous agents have demonstrated strong capabilities in reasoning, planning, and tool use, yet remain limited when tasks require sustained coordination across roles, tools, and environments. Multi-agent systems address this through structured collaboration among specialized agents, but tighter coordination also amplifies a less explored risk: errors can propagate across agents and interaction rounds, producing failures that are difficult to diagnose and rarely translate into structural self-improvement. Existing surveys cover individual agent capabilities, multi-agent collaboration, or agent self-evolution separately, leaving the causal dependencies among them unexamined. This survey provides a unified review organized around four causally linked stages, which we term the LIFE progression: Lay the capability foundation, Integrate agents through collaboration, Find faults through attribution, and Evolve through autonomous self-improvement. For each stage, we provide systematic taxonomies and formally characterize the dependencies between adjacent stages, revealing how each stage both depends on and constrains the next. Beyond synthesizing existing work, we identify open challenges at stage boundaries and propose a cross-stage research agenda for closed-loop multi-agent systems capable of continuously diagnosing failures, reorganizing structures, and refining agent behaviors, extending current coordination frameworks toward more self-organizing forms of collective intelligence. By bridging these previously fragmented research threads, this survey aims to offer both a systematic reference and a conceptual roadmap toward autonomous, self-improving multi-agent intelligence.
Engineering Breakdown
Plain English
This survey examines how errors propagate and compound across multiple LLM-based agents working together—a problem that gets worse the more agents coordinate. Rather than treating individual agent capability, multi-agent collaboration, and agent self-improvement as separate topics, the authors map causal dependencies between these three areas and propose the LIFE framework to organize how errors cascade through agent systems and whether agents can actually learn from those failures.
Key Engineering Insight
Error propagation in multi-agent systems isn't just about individual agent mistakes—it's about how errors compound across interaction rounds and agent roles, making root cause analysis nearly impossible without explicit failure attribution mechanisms built into the system architecture.
Why It Matters for Engineers
Today's production multi-agent systems (customer support chains, code generation pipelines, research assistants) fail silently at scale: you can't tell if Agent A gave Agent B bad data, Agent B misinterpreted correct data, or the system's coordination logic broke. Without understanding these causal chains and building debugging infrastructure, you can't diagnose failures or improve the system—you just patch symptoms.
Research Context
Prior work treated agent reasoning, team coordination, and learning independently. This survey connects the dots: showing that coordination amplifies individual weaknesses, and that without proper failure attribution, agents can't self-correct. It establishes LIFE as a framework that ties together collaboration patterns, diagnosis mechanisms, and self-evolution strategies—creating a foundation for building more reliable autonomous agent systems.
:::tip Subscribe Get weekly breakdowns of papers like this in AI Letters - the newsletter for engineers building production AI systems. :::