OneGen: Efficient One-Pass Unified Generation and Retrieval for LLMs
This addresses the need for seamless integration of retrieval and generation in practical applications, representing a novel method rather than an incremental improvement.
The paper tackles the limitation of LLMs in handling retrieval tasks by introducing OneGen, a framework that unifies generation and retrieval in a single forward pass, showing improved retrieval performance while preserving generative capabilities on tasks like RAG and Entity Linking.
Despite the recent advancements in Large Language Models (LLMs), which have significantly enhanced the generative capabilities for various NLP tasks, LLMs still face limitations in directly handling retrieval tasks. However, many practical applications demand the seamless integration of both retrieval and generation. This paper introduces a novel and efficient One-pass Generation and retrieval framework (OneGen), designed to improve LLMs' performance on tasks that require both generation and retrieval. The proposed framework bridges the traditionally separate training approaches for generation and retrieval by incorporating retrieval tokens generated autoregressively. This enables a single LLM to handle both tasks simultaneously in a unified forward pass. We conduct experiments on two distinct types of composite tasks, RAG and Entity Linking, to validate the pluggability, effectiveness, and efficiency of OneGen in training and inference. Furthermore, our results show that integrating generation and retrieval within the same context preserves the generative capabilities of LLMs while improving retrieval performance. To the best of our knowledge, OneGen is the first to enable LLMs to conduct vector retrieval during the generation.