AI Agent Systems: Architectures, Applications, and Evaluation

Системы ИИ-агентов: архитектуры, приложения и оценка
Bin Xu
2026-01-05

AI agent systemsagent architecturesagent evaluationmulti-agent coordinationtool use
AI agents -- systems that combine foundation models with reasoning, planning, memory, and tool use -- are rapidly becoming a practical interface between natural-language intent and real-world computation. This survey synthesizes the emerging landscape of AI agent architectures across: (i) deliberation and reasoning (e.g., chain-of-thought-style decomposition, self-reflection and verification, and constraint-aware decision making), (ii) planning and control (from reactive policies to hierarchical and multi-step planners), and (iii) tool calling and environment interaction (retrieval, code execution, APIs, and multimodal perception). We organize prior work into a unified taxonomy spanning agent components (policy/LLM core, memory, world models, planners, tool routers, and critics), orchestration patterns (single-agent vs.\ multi-agent; centralized vs.\ decentralized coordination), and deployment settings (offline analysis vs.\ online interactive assistance; safety-critical vs.\ open-ended tasks). We discuss key design trade-offs -- latency vs.\ accuracy, autonomy vs.\ controllability, and capability vs.\ reliability -- and highlight how evaluation is complicated by non-determinism, long-horizon credit assignment, tool and environment variability, and hidden costs such as retries and context growth. Finally, we summarize measurement and benchmarking practices (task suites, human preference and utility metrics, success under constraints, robustness and security) and identify open challenges including verification and guardrails for tool actions, scalable memory and context management, interpretability of agent decisions, and reproducible evaluation under realistic workloads.
1
AI agents combine foundation models with reasoning, planning, memory, and tool use to connect natural-language intent with real-world computation.
2
Agent architectures span deliberative reasoning, self-reflection and verification, hierarchical planning, tool routing, retrieval, code execution, APIs, and multimodal environment interaction.
3
Agent design involves fundamental trade-offs among latency and accuracy, autonomy and controllability, and capability and reliability.
4
Evaluation is complicated by nondeterminism, long-horizon credit assignment, variable tools and environments, retries, and context-growth costs.
5
Major open challenges include verifying tool actions, implementing guardrails, scaling memory and context management, interpreting agent decisions, and reproducing evaluations under realistic workloads.
6
The survey develops a unified taxonomy covering agent components, orchestration patterns, and deployment settings, including single- versus multi-agent systems and centralized versus decentralized coordination.

AI agent systems

architectures, applications, and evaluation, including reasoning, planning, tool use, coordination, reliability, and benchmarking

Publication Details
Publication Date
2026-01-05
Journal
Publisher
ISSN
Access Type
Author Information
Authors
Bin Xu
Explore further
Open the scid.ai AI chat with a ready-made request: it will find papers on a similar topic and help build a literature review.
Find similar papers in the chat →
Make a presentation
100%