about
Agentic RAG: Taxonomy, Architectures, Evaluation and Research Directions (arxiv.org)
2 points by teleforce 35 days ago | hide | past | pdf | discuss on HN

In plain words: They organized fast-growing AI search-and-answer systems that plan, remember, and call tools over many steps into one framework, treating each as a step-by-step decision process. The review found designs and tests are inconsistent, with risks like errors compounding and poisoned memory.

Abstract · SoK: Agentic Retrieval-Augmented Generation (RAG): Taxonomy, Architectures, Evaluation, and Research Directions

Retrieval-Augmented Generation (RAG) systems are increasingly evolving into agentic architectures where large language models autonomously coordinate multi-step reasoning, dynamic memory management, and iterative retrieval strategies. Despite rapid industrial adoption, current research lacks a systematic understanding of Agentic RAG as a sequential decision-making system, leading to highly fragmented architectures, inconsistent evaluation methodologies, and unresolved reliability risks. This Systematization of Knowledge (SoK) paper provides the first unified framework for understanding these autonomous systems. We formalize agentic retrieval-generation loops as finite-horizon partially observable Markov decision processes, explicitly modeling their control policies and state transitions. Building upon this formalization, we develop a comprehensive taxonomy and modular architectural decomposition that categorizes systems by their planning mechanisms, retrieval orchestration, memory paradigms, and tool-invocation behaviors. We further analyze the critical limitations of traditional static evaluation practices and identify severe systemic risks inherent to autonomous loops, including compounding hallucination propagation, memory poisoning, retrieval misalignment, and cascading tool-execution vulnerabilities. Finally, we outline key doctoral-scale research directions spanning stable adaptive retrieval, cost-aware orchestration, formal trajectory evaluation, and oversight mechanisms, providing a definitive roadmap for building reliable, controllable, and scalable agentic retrieval systems.

Saroj Mishra, Suman Niroula, Umesh Yadav, Dilip Thakur, Srijan Gyawali, Shiva Gaire
arXiv:2603.07379 · cs.AI, cs.CL, cs.CR, cs.IR · submitted Mar 7, 2026
abstract · pdf · html

add comment on HN