Stranger, Fan, or Peer? A Systematic Study on the Role of Interlocutor in Persona-Based Dialogue Generation

2026-08-28

AI Summary

研究通过控制对话双方在训练和推理时对彼此背景信息的可见性，探讨了基于角色的对话生成中角色特质表达与背景文本复制的问题。

Abstract

Persona-based dialogue systems are usually conditioned on speaker biography, but dialogues involve at least two participants, and who has access to whose biography can vary across training, inference, and evaluation. Prior work often neglected these aspects, obscuring mechanisms that only appear when biography visibility is toggled separately across training, inference, and evaluation, a three-stage factorisation that prior work has largely treated as a single factor. We study this factorisation on a dataset of dialogues paired with speaker's biographies, varying whether the target and interlocutor speakers see each other's biographies during training and inference, and using an LLM as a judge to perform author identification. We find that (i) training-time visibility, more than inference-time visibility, determines whether models express persona traits through dialogue or fall back on copying biographical text (a known problem/phenomenon in persona-based generation); (ii) models trained with interlocutor-biography visibility copy less target-biographical text than models trained without it, while changing visibility only at inference time has a less consistent effect; and (iii) under asymmetric disclosure, where only the interlocutor sees the target biography, target content leaks into interlocutor turns more often, and dialogues containing such traces are easier for the judge to identify, especially when interlocutor turns are visible. These results suggest that biography leakage into generated turns is an artefact of how interlocutor visibility is configured across training and inference, and separating the three stages is necessary.

Problem

• Persona-based Dialogue
• Biography Visibility
• Training and Inference
• Author Identification
• Biography Leakage

Innovation

• interlocutor visibility
• persona-based dialogue generation
• biography leakage
• training and inference stages