A comprehensive index of artificial intelligence and machine-learning research with AI-generated summaries, citation metrics, and direct links to papers and code.
Fei Wang, Xingchen Wan, Ruoxi Sun, et al.
Astute RAG adaptively elicits internal knowledge from LLMs and iteratively consolidates it with external sources to overcome imperfect retrieval in RAG.
Alexandria Leto, Cecilia Aguerrebere, I. Bhati, et al.
This paper investigates how retrieval accuracy affects RAG pipeline performance for QA tasks, finding that lowering search accuracy minimally impacts downstream quality while improving speed and memory.
Zhengren Wang, Jiayang Yu, Dongsheng Ma, et al.
RARE decouples knowledge storage from reasoning by externalizing domain knowledge to retrievable sources and internalizing reasoning patterns, enabling lightweight models to surpass GPT-4 and DeepSeek-R1 by ~20% accuracy.
Kaiwen Wei, Rui Shan, Dongsheng Zou, et al.
MIRAGE enhances test-time scaling for medical QA by combining multi-path parallel inference with structured knowledge graph retrieval to reduce error accumulation and improve traceability.
Orion Weller, Michael Boratko, Iftekhar Naim, et al.
This paper proves that single-vector embedding models have a fundamental dimension-dependent limit on the number of distinct top-k subsets they can represent, and shows this limitation arises even in simple realistic queries.
Zackary Rackauckas
Evaluates RAG-Fusion combining RAG and reciprocal rank fusion for product information retrieval, finding accurate answers but occasional off-topic responses.
Zongxi Li, Zijian Wang, Weiming Wang, et al.
A systematic survey of Retrieval-Augmented Generation (RAG) in education, covering workflow, retrievers, generation optimization, and applications.
Jakub Swacha, Michał Gracel
A survey of 47 papers on RAG chatbots in education, analyzing their character, target support, knowledge scope, LLM, and evaluation.
Genc Hoxha, Farid Melgani, Begüm Demir
Proposes a remote sensing image retrieval system that generates and exploits textual captions to describe high-level semantic content, improving retrieval accuracy.
Zheyang Xiong, Vasileios Papageorgiou, Kangwook Lee, et al.
Proposes finetuning LLMs on synthetic key-value retrieval data to improve long-context retrieval and reasoning without harming general benchmarks.
Mo Li, Songyang Zhang, Taolin Zhang, et al.
NeedleBench is a synthetic framework for evaluating LLMs' retrieval and reasoning in long contexts, revealing that reasoning models like Deepseek-R1 and o3 struggle with continuous retrieval in information-dense scenarios due to 'under-thinking'.
Dongyu Ru, Lin Qiu, Xiangkun Hu, et al.
RAGChecker is a fine-grained evaluation framework with diagnostic metrics for retrieval and generation modules in RAG systems, showing better correlation with human judgments.