GRAG: Generic Response-Augmented Generation Framework for Personalized Conversational Systems

2026-06-19

AI Summary

This work addresses the challenge of simultaneously achieving contextual grounding and personalization in dialogue systems under resource-constrained or privacy-sensitive settings. To this end, the authors propose GRAG, a novel framework that decouples content grounding from personalization for the first time. GRAG leverages offline responses generated by a general-purpose large language model as semantic and structural scaffolds to guide the efficient fine-tuning of a compact task-specific model. Departing from conventional end-to-end training paradigms, the framework supports both pre-fusion and post-fusion architectures. Extensive experiments demonstrate that GRAG significantly outperforms existing approaches across multiple personalized dialogue benchmarks, achieving up to a 47% relative improvement in ROUGE-2 and a 36% gain in BLEU scores.

Abstract

Deploying highly capable personalized conversational agents in resource-constrained or privacy-sensitive environments remains a significant challenge. We identify a fundamental bottleneck in the existing approaches: current training paradigms treat personalization and grounding as a single monolithic learning problem. Under these paradigms, language models are forced to simultaneously address what to say (content grounding) and how to say it in a user-specific way (personalization), which introduces significant computational and optimization challenges. Consequently, contextual grounding is often sacrificed for persona adherence, or vice versa, resulting in responses that are either weakly grounded in the conversational history or insufficiently personalized. In this work, we propose the Generic Response-Augmented Generation (GRAG) framework that decouples these competing objectives by leveraging offline, generic responses from high-capacity, general-purpose LLMs as a semantic and structural scaffold to guide the fine-tuning of smaller, task-specialized models seamlessly in resource-limited environments. By decoupling the content grounding from personalization, GRAG allows the model to focus exclusively on persona injection while remaining firmly anchored to the conversational context. We instantiate the GRAG in two post- and pre-fusion-based architectural variants and evaluate them on multiple benchmark conversational datasets that cover diverse personalization structures. Our results demonstrate that GRAG significantly outperforms state-of-the-art methods that do not use auxiliary scaffolding, yielding up to 47% improvements in ROUGE-2 and 36% in BLEU scores. Ultimately, GRAG offers a generalizable blueprint for building grounding-aware personalized conversational systems in resource-limited environments.

Problem

• personalized conversational systems
• contextual grounding
• personalization
• resource-constrained environments
• language models

Innovation

• response-augmented generation
• personalized conversational systems
• grounding-personalization decoupling
• resource-constrained LLM fine-tuning
• generic response scaffolding