DeepResearch: A Survey of LLM-based Research Agents

Large Language Model Technologies: Progress, Problems and Prospects

Large language models (LLMs) are built on the classic Transformer architecture and have become a core driving force for the rapid development of modern artificial intelligence. This paper presents a systematic review of LLMs, elaborating on their fundamental working principles, mainstream open-source models, effective lightweight optimization methods, retrieval-augmented generation frameworks and key human-value-aligned technologies. Nowadays, LLMs have been widely applied in practice. Typical scenarios include intelligent text generation, professional knowledge-based question answering and automated code generation, delivering remarkable value to both industries and academia. However, their large-scale industrial application is still restricted by multiple challenges. The major issues involve content hallucination, poor model interpretability, excessive computing resource consumption, potential ethical risks and unsatisfactory multimodal integration capability. This paper also forecasts the future development directions of LLMs, such as lightweight deployment on edge devices, safety-focused human value alignment, in-depth cross-modal fusion and customized large models for vertical industries. Additionally, it collects a number of representative cases, which can offer solid references and practical guidance for relevant researchers and engineering practitioners to carry out further studies.

The Versatility of Large Language Models: A Comprehensive Review and Structured Survey of Architectures, Applications, Challenges, and Future Trajectories

This survey reviews the evolution of language models from early statistical approaches to modern Transformer-based architectures and summarizes key developments, including attention mechanisms, scaling laws, alignment techniques, and efficient inference methods.

Adesso Intelligent Agent - An Addon Ecosystem for Extending and Empowering Rag Systems

A comparative review of modern large language model paradigms: GPT-4, BERT, Gemini, and DeepSeek

Comparison of GPT-4, BERT (bidirectional encoder representations from transformers), Gemini, and DeepSeek large language models (LLM), focusing on architectures, training methodologies, and real-world applications reveals GPT-4 excels in natural language generation and complex reasoning, supporting up to 128K tokens with moderate latency and higher costs making it effective for conversational artificial intelligence (AI).

SmartRGPT: An NLP-Based Conversational Research Support System Designed Using Open-Source RAG and LLM

This case study research aims to focus on designing a conversational Research Support System (RSS) named SmartRGPT using open-source Retrieval Augmented Generation (RAG) and a Large Language Model (LLM) to enhance and address the challenges faced by researchers while using the existing Research Support Services called SmartRLibrary (initiated by and applied in B C Roy Memorial Library, alternatively, IIM Calcutta Library). It also addresses the limitations of traditional keyword-based search and the hallucination issues of standalone LLM. The prototype has been designed using several open-source software components, including the RAG pipeline, LangChain, the ChromaDB vector database, and the Llama-3 (70-billion-parameter model). A curated set of over 250 datasets was collected, preprocessed, and ingested using Wget (WarcGPT framework) for preparing the knowledge base. The prototype was tested and evaluated using real-world queries. Based on internal review and initial observations of the authors on the generated responses, in the majority of tested cases, the findings demonstrate that the proposed system generated accurate, context‑aware responses without hallucinations. It has responded to short and long-range queries based on its ingested knowledge bases, citing the sources as references. The findings further indicate that the proposed system has the potential to provide 24/7 personalised research assistance, reduce repetitive library workload, and enable the library to provide more advanced services if applied after rigorous evaluation in larger populations. Its cost-effective open-source architecture also offers libraries with limited budgets an independent and customisable alternative to vendor-dependent solutions, thereby contributing to the advancement of the Library and Information Science (LIS) domain.

Towards Conversational Dataset Retrieval: A Survey

A novel, layered conceptual framework is introduced that organizes research in CDR across four key dimensions: User Layer, System Layer, Data Layer, and Evaluation Layer and identifies core challenges in CDR, including the lack of standardized evaluation benchmarks and limited support for ambiguous or evolving user intent.