IntrAgent: An LLM Agent for Content-Grounded Information Retrieval through Literature Review

2026/04/23 by Fengbo Ma, Zixin Rao, Xiaoting Li +5 · 1 voice
Biochemistry, Genetics and Molecular Biology · Computer Science · #Biomedical Text Mining and Ontologies #Information Retrieval and Search Behavior #Topic Modeling #cs.AI #cs.IR #cs.LG

paper · pdf · doi:10.48550/arxiv.2604.22861

openalex publication_date 2026/04/23 · arxiv published 2026/04/23 · arxiv updated 2026/04/23 · openalex created_date 2026/04/29 · openalex updated_date 2026/07/28

Abstract

Scientific research relies on accurate information retrieval from literature to support analytical decisions. In this work, we introduce a new task, INformation reTRieval through literAture reVIEW (IntraView), which aims to automate fine-grained information retrieval faithfully grounded in the provided content in response to research-driven queries, and propose IntrAgent, an LLM-based agent that addresses this challenging task. In particular, IntrAgent is designed to mimic human behaviors when reading literature for information retrieval -- identifying relevant sections and then iteratively extracting key details to refine the retrieved information. It follows a two-stage pipeline: a Section Ranking stage that prioritizes relevant literature sections through structural-knowledge-enabled reasoning, and an Iterative Reading stage that continuously extracts details and synthesizes them into concise, contextually grounded answers. To support rigorous evaluation, we introduce IntraBench, a new benchmark consisting of 315 test instances built from expert-authored questions paired with literature spanning five STEM domains. Across seven backbone LLMs, IntrAgent achieves on average 13.2% higher cross-domain accuracy than state-of-the-art RAG and research-agent baselines.

Citations

• Open-Source Agentic Hybrid RAG Framework for Scientific Literature Review
• ChemAgent: Self-updating Library in Large Language Models Improves Chemical Reasoning
• Agent Laboratory: Using LLM Agents as Research Assistants
• Crafting Personalized Agents through Retrieval-Augmented Generation on Editable Memory Graphs
• Language agents achieve superhuman synthesis of scientific knowledge
• HoneyComb: A Flexible LLM-Based Agent System for Materials Science
• The AI Scientist: Towards Fully Automated Open-Ended Scientific Discovery
• Retrieval-Augmented Generation for Natural Language Processing: A Survey
• LongRAG: Enhancing Retrieval-Augmented Generation with Long-context LLMs
• A Comprehensive Survey of Scientific Large Language Models and Their Applications in Scientific Discovery
• M-RAG: Reinforcing Large Language Model Performance through Retrieval-Augmented Generation with Multiple Partitions
• Efficient Encoder-Decoder Transformer Decoding for Decomposable Tasks
• Enhancing LLM Factual Accuracy with RAG to Counter Hallucinations: A Case Study on Domain-Specific Queries in Private Knowledge-Bases
• HaluEval-Wild: Evaluating Hallucinations of Language Models in the Wild
• Assisting in Writing Wikipedia-like Articles From Scratch with Large Language Models
• OlympiadBench: A Challenging Benchmark for Promoting AGI with Olympiad-Level Bilingual Multimodal Scientific Problems
• LitLLM: A Toolkit for Scientific Literature Review
• PaperQA: Retrieval-Augmented Generative Agent for Scientific Research
• GPQA: A Graduate-Level Google-Proof Q&A Benchmark
• Agent Lumos: Unified and Modular Training for Open-Source Language Agents
• MathVista: Evaluating Mathematical Reasoning of Foundation Models in Visual Contexts
• ChemCrow: Augmenting large-language models with chemistry tools
• Learn to Explain: Multimodal Reasoning via Thought Chains for Science Question Answering
• A Dataset of Information-Seeking Questions and Answers Anchored in\n Research Papers
• PRAG: Paninian Retrieval-Augmented Generation for Safety-Critical Medical Question Answering
• Retrieval-Augmented Generation for Knowledge-Intensive NLP Tasks
• Mathematical modeling of COVID-19 transmission dynamics with a case study of Wuhan
• HybridQA: A Dataset of Multi-Hop Question Answering over Tabular and Textual Data
• MiniLM: Deep Self-Attention Distillation for Task-Agnostic Compression of Pre-Trained Transformers
• PubMedQA: A Dataset for Biomedical Research Question Answering
• CoQA: A Conversational Question Answering Challenge
• Can a Suit of Armor Conduct Electricity? A New Dataset for Open Book Question Answering
• Know What You Don't Know: Unanswerable Questions for SQuAD
• Crowdsourcing Multiple Choice Science Questions
• TriviaQA: A Large Scale Distantly Supervised Challenge Dataset for\n Reading Comprehension
• Reading Wikipedia to Answer Open-Domain Questions
• The magical number seven, plus or minus two: Some limits on our capacity for processing information.

Discussions

• Conference Paper (preprint): "IntrAgent: An LLM Agent for Content-Grounded Information Retrieval through Literature Review arxiv.org/abs/2604.22861 #AgenticAI #AI #litreviews #LLMs [bsky, 2 points, 2 comments]