Modeling the One-to-Many Property in Open-Domain Dialogue with LLMs

TL;DR

This paper models the one-to-many property in open-domain dialogue using a two-stage approach with LLMs, improving response diversity and quality by leveraging a new dataset and novel strategies.

Contribution

It introduces o2mDial, a new dialogue corpus capturing multiple plausible responses, and proposes a two-stage framework with in-context learning and instruction tuning for better response diversity.

Findings

Enhanced response diversity in smaller LLMs

Response quality improved by up to 90%

Closer performance to larger models

Abstract

Open-domain Dialogue (OD) exhibits a one-to-many (o2m) property, whereby multiple appropriate responses exist for a single dialogue context. Despite prior research showing that modeling this property boosts response diversity, most modern LLM-based dialogue agents do not explicitly do so. In this work, we model the o2m property of OD in LLMs by decomposing OD generation into two key tasks: Multi-Response Generation (MRG) and Preference-based Selection (PS), which entail generating a set of n semantically and lexically diverse high-quality responses for a given dialogue context, followed by selecting a single response based on human preference, respectively. To facilitate MRG and PS, we introduce o2mDial, a dialogue corpus explicitly designed to capture the o2m property by featuring multiple plausible responses for each context. Leveraging o2mDial, we propose new in-context learning and instruction-tuning strategies, as well as novel evaluation metrics for MRG, alongside a model-based approach for PS. Empirical results demonstrate that applying the proposed two-stage framework to smaller LLMs for OD generation enhances overall response diversity while maintaining contextual coherence, improving response quality by up to 90%, bringing them closer to the performance of larger models.