While large language models (LLMs) have achieved state-of-the-art performance on a wide range of medical question answering (QA) tasks, they still face challenges with hallucinations and outdated knowledge. Retrieval-augmented generation (RAG) is a promising solution and has been widely adopted. However, a RAG system can involve multiple flexible components, and there is a lack of best practices regarding the optimal RAG setting for various medical purposes. To systematically evaluate such systems, we propose the Medical Information Retrieval-Augmented Generation Evaluation (MIRAGE), a first-of-its-kind benchmark including 7,663 questions from five medical QA datasets. Using MIRAGE, we conducted large-scale experiments with over 1.8 trillion prompt tokens on 41 combinations of different corpora, retrievers, and backbone LLMs through the MedRAG toolkit introduced in this work. Overall, MedRAG improves the accuracy of six different LLMs by up to 18% over chain-of-thought prompting, elevating the performance of GPT-3.5 and Mixtral to GPT-4-level. Our results show that the combination of various medical corpora and retrievers achieves the best performance. In addition, we discovered a log-linear scaling property and the "lost-in-the-middle" effects in medical RAG. We believe our comprehensive evaluations can serve as practical guidelines for implementing RAG systems for medicine.
Retrieval-augmented generation framework can address the limitations of large language models by enabling real-time knowledge updates for more accurate answers. An efficient way in the training phase of retrieval-augmented models is attention distillation, which uses attention scores as a supervision signal instead of manually annotated query-document pairs. Despite its growing popularity, the detailed mechanisms behind the success of attention distillation remain unexplored, particularly the specific patterns it leverages to benefit training. In this paper, we address this gap by conducting a comprehensive review of attention distillation workflow and identifying key factors influencing the learning quality of retrieval-augmented language models. We further propose indicators for optimizing models' training methods and avoiding ineffective training.
Retrieval-augmented generation have become central in natural language processing due to their efficacy in generating factual content. While traditional methods employ single-time retrieval, more recent approaches have shifted towards multi-time retrieval for multi-hop reasoning tasks. However, these strategies are bound by predefined reasoning steps, potentially leading to inaccuracies in response generation. This paper introduces MetaRAG, an approach that combines the retrieval-augmented generation process with metacognition. Drawing from cognitive psychology, metacognition allows an entity to self-reflect and critically evaluate its cognitive processes. By integrating this, MetaRAG enables the model to monitor, evaluate, and plan its response strategies, enhancing its introspective reasoning abilities. Through a three-step metacognitive regulation pipeline, the model can identify inadequacies in initial cognitive responses and fixes them. Empirical evaluations show that MetaRAG significantly outperforms existing methods.
Dense passage retrieval (DPR) is the first step in the retrieval augmented generation (RAG) paradigm for improving the performance of large language models (LLM). DPR fine-tunes pre-trained networks to enhance the alignment of the embeddings between queries and relevant textual data. A deeper understanding of DPR fine-tuning will be required to fundamentally unlock the full potential of this approach. In this work, we explore DPR-trained models mechanistically by using a combination of probing, layer activation analysis, and model editing. Our experiments show that DPR training decentralizes how knowledge is stored in the network, creating multiple access pathways to the same information. We also uncover a limitation in this training style: the internal knowledge of the pre-trained model bounds what the retrieval model can retrieve. These findings suggest a few possible directions for dense retrieval: (1) expose the DPR training process to more knowledge so more can be decentralized, (2) inject facts as decentralized representations, (3) model and incorporate knowledge uncertainty in the retrieval process, and (4) directly map internal model knowledge to a knowledge base.
Existing work on Temporal Question Answering (TQA) has predominantly focused on questions anchored to specific timestamps or events (e.g. "Who was the US president in 1970?"). Little work has studied questions whose temporal context is relative to the present time (e.g. "Who was the previous US president?"). We refer to this problem as Present-Anchored Temporal QA (PATQA). PATQA poses unique challenges: (1) large language models (LLMs) may have outdated knowledge, (2) complex temporal relationships (e.g. 'before', 'previous') are hard to reason, (3) multi-hop reasoning may be required, and (4) the gold answers of benchmarks must be continuously updated. To address these challenges, we introduce the PAT-Questions benchmark, which includes single and multi-hop temporal questions. The answers in PAT-Questions can be automatically refreshed by re-running SPARQL queries on a knowledge graph, if available. We evaluate several state-of-the-art LLMs and a SOTA temporal reasoning model (TEMPREASON-T5) on PAT-Questions through direct prompting and retrieval-augmented generation (RAG). The results highlight the limitations of existing solutions in PATQA and motivate the need for new methods to improve PATQA reasoning capabilities.
Large language models (LLMs) are transforming the way information is retrieved with vast amounts of knowledge being summarized and presented via natural language conversations. Yet, LLMs are prone to highlight the most frequently seen pieces of information from the training set and to neglect the rare ones. In the field of biomedical research, latest discoveries are key to academic and industrial actors and are obscured by the abundance of an ever-increasing literature corpus (the information overload problem). Surfacing new associations between biomedical entities, e.g., drugs, genes, diseases, with LLMs becomes a challenge of capturing the long-tail knowledge of the biomedical scientific production. To overcome this challenge, Retrieval Augmented Generation (RAG) has been proposed to alleviate some of the shortcomings of LLMs by augmenting the prompts with context retrieved from external datasets. RAG methods typically select the context via maximum similarity search over text embeddings. In this study, we show that RAG methods leave out a significant proportion of relevant information due to clusters of over-represented concepts in the biomedical literature. We introduce a novel information-retrieval method that leverages a knowledge graph to downsample these clusters and mitigate the information overload problem. Its retrieval performance is about twice better than embedding similarity alternatives on both precision and recall. Finally, we demonstrate that both embedding similarity and knowledge graph retrieval methods can be advantageously combined into a hybrid model that outperforms both, enabling potential improvements to biomedical question-answering models.
Federated search systems aggregate results from multiple search engines, selecting appropriate sources to enhance result quality and align with user intent. With the increasing uptake of Retrieval-Augmented Generation (RAG) pipelines, federated search can play a pivotal role in sourcing relevant information across heterogeneous data sources to generate informed responses. However, existing datasets, such as those developed in the past TREC FedWeb tracks, predate the RAG paradigm shift and lack representation of modern information retrieval challenges. To bridge this gap, we present FeB4RAG, a novel dataset specifically designed for federated search within RAG frameworks. This dataset, derived from 16 sub-collections of the widely used \beir benchmarking collection, includes 790 information requests (akin to conversational queries) tailored for chatbot applications, along with top results returned by each resource and associated LLM-derived relevance judgements. Additionally, to support the need for this collection, we demonstrate the impact on response generation of a high quality federated search system for RAG compared to a naive approach to federated search. We do so by comparing answers generated through the RAG pipeline through a qualitative side-by-side comparison. Our collection fosters and supports the development and evaluation of new federated search methods, especially in the context of RAG pipelines.
Recently the retrieval-augmented generation (RAG) has been successfully applied in code generation. However, existing pipelines for retrieval-augmented code generation (RACG) employ static knowledge bases with a single source, limiting the adaptation capabilities of Large Language Models (LLMs) to domains they have insufficient knowledge of. In this work, we develop a novel pipeline, EVOR, that employs the synchronous evolution of both queries and diverse knowledge bases. On two realistic settings where the external knowledge is required to solve code generation tasks, we compile four new datasets associated with frequently updated libraries and long-tail programming languages, named EVOR-BENCH. Extensive experiments demonstrate that EVOR achieves two to four times of execution accuracy compared to other methods such as Reflexion (Shinn et al., 2024), DocPrompting (Zhou et al., 2023), etc. We demonstrate that EVOR is flexible and can be easily combined with them to achieve further improvement. Further analysis reveals that EVOR benefits from the synchronous evolution of queries and documents and the diverse information sources in the knowledge base. We hope that our studies will inspire more insights into the design of advanced RACG pipelines in future research. Our model, code, and data are available at https://arks-codegen.github.io.
대규모 언어 모델(LLM)의 환각 문제를 해결하기 위한 효과적인 방법으로 **Retrieval Augmented Generation (RAG)**이 주목받고 있다. RAG의 검색 단계에서는 사전 학습된 임베딩 모델이 쿼리와 패시지를 벡터로 변환하여 의미를 포착하는 역할을 한다. 그러나 이러한 표준 임베딩 모델은 특정 도메인 지식에 대한 적용 시 최적의 성능을 발휘하지 못하는 경우가 많아, 파인튜닝이 필요하게 된다. 본 연구에서는 임베딩이 블랙박스 모델에서만 제공되는 상황을 다루며, 이를 해결하기 위한 새로운 접근법인 **모델 보강 파인튜닝 (Mafin)**을 제안한다.
Mafin은 학습 가능한 임베딩 모델을 통해 블랙박스 임베딩 모델을 보강하여 파인튜닝하는 방법론이다. 이 접근법은 소규모 보강 모델의 학습만으로 블랙박스 임베딩의 성능을 크게 향상시킬 수 있음을 보여준다. 연구 결과는 레이블이 있는 데이터셋과 레이블이 없는 데이터셋 모두에서 검증되었으며, Mafin의 적용 가능성과 효율성을 입증하였다.
Mafin의 핵심은 쿼리와 후보 패시지 간의 관련성을 평가하는 데 있다. 이를 위해 관련성 점수를 정의하고, KL 발산을 최소화하는 방법을 통해 모델 파라미터를 학습한다. 또한, 계산 효율성을 높이기 위해 상위 1위 근사 방법을 도입하여, 패시지 수가 많을 경우에도 효율적인 학습이 가능하도록 설계되었다.
이 연구는 블랙박스 임베딩 모델의 성능을 향상시키는 혁신적인 방법론을 제시하며, 특정 도메인 지식에 대한 적용 가능성을 높인다. Mafin은 향후 다양한 분야에서의 정보 검색 및 생성 작업에 기여할 수 있는 잠재력을 지니고 있다.
The construction industry is a vital sector of the global economy, but it faces many productivity challenges in various processes, such as design, planning, procurement, inspection, and maintenance. Generative artificial intelligence (AI), which can create novel and realistic data or content, such as text, image, video, or code, based on some input or prior knowledge, offers innovative and disruptive solutions to address these challenges. However, there is a gap in the literature on the current state, opportunities, and challenges of generative AI in the construction industry. This study aims to fill this gap by providing a state-of-the-art analysis of generative AI in construction, with three objectives: (1) to review and categorize the existing and emerging generative AI opportunities and challenges in the construction industry; (2) to propose a framework for construction firms to build customized generative AI solutions using their own data, comprising steps such as data collection, dataset curation, training custom large language model (LLM), model evaluation, and deployment; and (3) to demonstrate the framework via a case study of developing a generative model for querying contract documents. The results show that retrieval augmented generation (RAG) improves the baseline LLM by 5.2, 9.4, and 4.8% in terms of quality, relevance, and reproducibility. This study provides academics and construction professionals with a comprehensive analysis and practical framework to guide the adoption of generative AI techniques to enhance productivity, quality, safety, and sustainability across the construction industry.
All text-based language problems can be reduced to either generation or embedding. Current models only perform well at one or the other. We introduce generative representational instruction tuning (GRIT) whereby a large language model is trained to handle both generative and embedding tasks by distinguishing between them through instructions. Compared to other open models, our resulting GritLM 7B sets a new state of the art on the Massive Text Embedding Benchmark (MTEB) and outperforms all models up to its size on a range of generative tasks. By scaling up further, GritLM 8x7B outperforms all open generative language models that we tried while still being among the best embedding models. Notably, we find that GRIT matches training on only generative or embedding data, thus we can unify both at no performance loss. Among other benefits, the unification via GRIT speeds up Retrieval-Augmented Generation (RAG) by > 60% for long documents, by no longer requiring separate retrieval and generation models. Models, code, etc. are freely available at https://github.com/ContextualAI/gritlm.
This paper presents a novel Chunking-Free In-Context (CFIC) retrieval approach, specifically tailored for Retrieval-Augmented Generation (RAG) systems. Traditional RAG systems often struggle with grounding responses using precise evidence text due to the challenges of processing lengthy documents and filtering out irrelevant content. Commonly employed solutions, such as document chunking and adapting language models to handle longer contexts, have their limitations. These methods either disrupt the semantic coherence of the text or fail to effectively address the issues of noise and inaccuracy in evidence retrieval. CFIC addresses these challenges by circumventing the conventional chunking process. It utilizes the encoded hidden states of documents for in-context retrieval, employing auto-aggressive decoding to accurately identify the specific evidence text required for user queries, eliminating the need for chunking. CFIC is further enhanced by incorporating two decoding strategies, namely Constrained Sentence Prefix Decoding and Skip Decoding. These strategies not only improve the efficiency of the retrieval process but also ensure that the fidelity of the generated grounding text evidence is maintained. Our evaluations of CFIC on a range of open QA datasets demonstrate its superiority in retrieving relevant and accurate evidence, offering a significant improvement over traditional methods. By doing away with the need for document chunking, CFIC presents a more streamlined, effective, and efficient retrieval solution, making it a valuable advancement in the field of RAG systems.
The robustness of large language models (LLMs) becomes increasingly important as their use rapidly grows in a wide range of domains. Retrieval-Augmented Generation (RAG) is considered as a means to improve the trustworthiness of text generation from LLMs. However, how the outputs from RAG-based LLMs are affected by slightly different inputs is not well studied. In this work, we find that the insertion of even a short prefix to the prompt leads to the generation of outputs far away from factually correct answers. We systematically evaluate the effect of such prefixes on RAG by introducing a novel optimization technique called Gradient Guided Prompt Perturbation (GGPP). GGPP achieves a high success rate in steering outputs of RAG-based LLMs to targeted wrong answers. It can also cope with instructions in the prompts requesting to ignore irrelevant context. We also exploit LLMs' neuron activation difference between prompts with and without GGPP perturbations to give a method that improves the robustness of RAG-based LLMs through a highly effective detector trained on neuron activation triggered by GGPP generated prompts. Our evaluation on open-sourced LLMs demonstrates the effectiveness of our methods.
Large language models (LLMs) have achieved remarkable success due to their exceptional generative capabilities. Despite their success, they also have inherent limitations such as a lack of up-to-date knowledge and hallucination. Retrieval-Augmented Generation (RAG) is a state-of-the-art technique to mitigate these limitations. The key idea of RAG is to ground the answer generation of an LLM on external knowledge retrieved from a knowledge database. Existing studies mainly focus on improving the accuracy or efficiency of RAG, leaving its security largely unexplored. We aim to bridge the gap in this work. We find that the knowledge database in a RAG system introduces a new and practical attack surface. Based on this attack surface, we propose PoisonedRAG, the first knowledge corruption attack to RAG, where an attacker could inject a few malicious texts into the knowledge database of a RAG system to induce an LLM to generate an attacker-chosen target answer for an attacker-chosen target question. We formulate knowledge corruption attacks as an optimization problem, whose solution is a set of malicious texts. Depending on the background knowledge (e.g., black-box and white-box settings) of an attacker on a RAG system, we propose two solutions to solve the optimization problem, respectively. Our results show PoisonedRAG could achieve a 90% attack success rate when injecting five malicious texts for each target question into a knowledge database with millions of texts. We also evaluate several defenses and our results show they are insufficient to defend against PoisonedRAG, highlighting the need for new defenses.
Given a graph with textual attributes, we enable users to `chat with their graph': that is, to ask questions about the graph using a conversational interface. In response to a user's questions, our method provides textual replies and highlights the relevant parts of the graph. While existing works integrate large language models (LLMs) and graph neural networks (GNNs) in various ways, they mostly focus on either conventional graph tasks (such as node, edge, and graph classification), or on answering simple graph queries on small or synthetic graphs. In contrast, we develop a flexible question-answering framework targeting real-world textual graphs, applicable to multiple applications including scene graph understanding, common sense reasoning, and knowledge graph reasoning. Toward this goal, we first develop a Graph Question Answering (GraphQA) benchmark with data collected from different tasks. Then, we propose our G-Retriever method, introducing the first retrieval-augmented generation (RAG) approach for general textual graphs, which can be fine-tuned to enhance graph understanding via soft prompting. To resist hallucination and to allow for textual graphs that greatly exceed the LLM's context window size, G-Retriever performs RAG over a graph by formulating this task as a Prize-Collecting Steiner Tree optimization problem. Empirical evaluations show that our method outperforms baselines on textual graph tasks from multiple domains, scales well with larger graph sizes, and mitigates hallucination.~\footnote{Our codes and datasets are available at: \url{https://github.com/XiaoxinHe/G-Retriever}}
The integration of multimodal Electronic Health Records (EHR) data has significantly improved clinical predictive capabilities. Leveraging clinical notes and multivariate time-series EHR, existing models often lack the medical context relevent to clinical tasks, prompting the incorporation of external knowledge, particularly from the knowledge graph (KG). Previous approaches with KG knowledge have primarily focused on structured knowledge extraction, neglecting unstructured data modalities and semantic high dimensional medical knowledge. In response, we propose REALM, a Retrieval-Augmented Generation (RAG) driven framework to enhance multimodal EHR representations that address these limitations. Firstly, we apply Large Language Model (LLM) to encode long context clinical notes and GRU model to encode time-series EHR data. Secondly, we prompt LLM to extract task-relevant medical entities and match entities in professionally labeled external knowledge graph (PrimeKG) with corresponding medical knowledge. By matching and aligning with clinical standards, our framework eliminates hallucinations and ensures consistency. Lastly, we propose an adaptive multimodal fusion network to integrate extracted knowledge with multimodal EHR data. Our extensive experiments on MIMIC-III mortality and readmission tasks showcase the superior performance of our REALM framework over baselines, emphasizing the effectiveness of each module. REALM framework contributes to refining the use of multimodal EHR data in healthcare and bridging the gap with nuanced medical context essential for informed clinical predictions.
Integrating large language models (LLMs) with knowledge graphs derived from domain-specific data represents an important advancement towards more powerful and factual reasoning. As these models grow more capable, it is crucial to enable them to perform multi-step inferences over real-world knowledge graphs while minimizing hallucination. While large language models excel at conversation and text generation, their ability to reason over domain-specialized graphs of interconnected entities remains limited. For example, can we query a LLM to identify the optimal contact in a professional network for a specific goal, based on relationships and attributes in a private database? The answer is no--such capabilities lie beyond current methods. However, this question underscores a critical technical gap that must be addressed. Many high-value applications in areas such as science, security, and e-commerce rely on proprietary knowledge graphs encoding unique structures, relationships, and logical constraints. We introduce a fine-tuning framework for developing Graph-aligned LAnguage Models (GLaM) that transforms a knowledge graph into an alternate text representation with labeled question-answer pairs. We demonstrate that grounding the models in specific graph-based knowledge expands the models' capacity for structure-based reasoning. Our methodology leverages the large-language model's generative capabilities to create the dataset and proposes an efficient alternate to retrieval-augmented generation styled methods.
Large Language Models (LLMs) have demonstrated the strong potential to assist both clinicians and the general public with their extensive medical knowledge. However, their application in healthcare is constrained due to concerns about the privacy of data used in training, which prevents the integration of private and personal information because of security and ethical issues. Moreover, if their capabilities can be enhanced with information retrieval to access up-to-date knowledge, the current integration of LLMs with Information retrieval lacks robustness to imperfect retrieval, which can hinder their effectiveness and even reduce overall performance. In this work, we address this challenge by introducing the Retrieval-Augmented Thought Process (RATP). Given access to external knowledge, RATP formulates the thought generation of LLMs as a multiple-step decision process. To optimise such a thought process, RATP leverages Monte-Carlo Tree Search and learns a proxy reward function that permits cost-efficient inference. On a private dataset of electronic medical records, deliberately excluded from any LLM training set, RATP achieves 35% additional accuracy compared to in-context retrieval-augmented generation for the question-answering task.
Large Language Models (LLMs) are increasingly used across various domains, from software development to cyber threat intelligence. Understanding all the different fields of cybersecurity, which includes topics such as cryptography, reverse engineering, and risk assessment, poses a challenge even for human experts. To accurately test the general knowledge of LLMs in cybersecurity, the research community needs a diverse, accurate, and up-to-date dataset. To address this gap, we present CyberMetric-80, CyberMetric-500, CyberMetric-2000, and CyberMetric-10000, which are multiple-choice Q&A benchmark datasets comprising 80, 500, 2000, and 10,000 questions respectively. By utilizing GPT-3.5 and Retrieval-Augmented Generation (RAG), we collected documents, including NIST standards, research papers, publicly accessible books, RFCs, and other publications in the cybersecurity domain, to generate questions, each with four possible answers. The results underwent several rounds of error checking and refinement. Human experts invested over 200 hours validating the questions and solutions to ensure their accuracy and relevance, and to filter out any questions unrelated to cybersecurity. We have evaluated and compared 25 state-of-the-art LLM models on the CyberMetric datasets. In addition to our primary goal of evaluating LLMs, we involved 30 human participants to solve CyberMetric-80 in a closed-book scenario. The results can serve as a reference for comparing the general cybersecurity knowledge of humans and LLMs. The findings revealed that GPT-4o, GPT-4-turbo, Mixtral-8x7B-Instruct, Falcon-180B-Chat, and GEMINI-pro 1.0 were the best-performing LLMs. Additionally, the top LLMs were more accurate than humans on CyberMetric-80, although highly experienced human experts still outperformed small models such as Llama-3-8B, Phi-2 or Gemma-7b.
Large Language Models (LLM) have shown remarkable language capabilities fueling attempts to integrate them into applications across a wide range of domains. An important application area is question answering over private enterprise documents where the main considerations are data security, which necessitates applications that can be deployed on-prem, limited computational resources and the need for a robust application that correctly responds to queries. Retrieval-Augmented Generation (RAG) has emerged as the most prominent framework for building LLM-based applications. While building a RAG is relatively straightforward, making it robust and a reliable application requires extensive customization and relatively deep knowledge of the application domain. We share our experiences building and deploying an LLM application for question answering over private organizational documents. Our application combines the use of RAG with a finetuned open-source LLM. Additionally, our system, which we call Tree-RAG (T-RAG), uses a tree structure to represent entity hierarchies within the organization. This is used to generate a textual description to augment the context when responding to user queries pertaining to entities within the organization's hierarchy. Our evaluations, including a Needle in a Haystack test, show that this combination performs better than a simple RAG or finetuning implementation. Finally, we share some lessons learned based on our experiences building an LLM application for real-world use.