CantTalkAboutThis: Aligning Language Models to Stay on Topic in Dialogues
Makesh Narsimhan Sreedhar, Traian Rebedea, Shaona Ghosh, Jiaqi Zeng, Christopher Parisien
Abstract
Recent advancements in instruction-tuning datasets have predominantly focused on specific tasks like mathematical or logical reasoning. There has been a notable gap in data designed for aligning language models to maintain topic relevance in conversations - a critical aspect for deploying chatbots to production. We introduce the CantTalkAboutThis dataset to help language models remain focused on the subject at hand during task-oriented interactions. It consists of synthetic dialogues on a wide range of conversation topics from different domains. These dialogues are interspersed with distractor turns that intentionally divert the chatbot from the predefined topic. Fine-tuning language models on this dataset helps make them resilient to deviating from the assigned role and improves their ability to maintain topical coherence compared to general-purpose instruction-tuned LLMs like gpt-4-turbo and Mixtral-Instruct. Additionally, preliminary observations suggest that training models on this dataset also enhance their performance on fine-grained instruction following tasks, including safety alignment.- Anthology ID:
- 2024.findings-emnlp.713
- Volume:
- Findings of the Association for Computational Linguistics: EMNLP 2024
- Month:
- November
- Year:
- 2024
- Address:
- Miami, Florida, USA
- Editors:
- Yaser Al-Onaizan, Mohit Bansal, Yun-Nung Chen
- Venue:
- Findings
- SIG:
- Publisher:
- Association for Computational Linguistics
- Note:
- Pages:
- 12232–12252
- Language:
- URL:
- https://aclanthology.org/2024.findings-emnlp.713
- DOI:
- 10.18653/v1/2024.findings-emnlp.713
- Cite (ACL):
- Makesh Narsimhan Sreedhar, Traian Rebedea, Shaona Ghosh, Jiaqi Zeng, and Christopher Parisien. 2024. CantTalkAboutThis: Aligning Language Models to Stay on Topic in Dialogues. In Findings of the Association for Computational Linguistics: EMNLP 2024, pages 12232–12252, Miami, Florida, USA. Association for Computational Linguistics.
- Cite (Informal):
- CantTalkAboutThis: Aligning Language Models to Stay on Topic in Dialogues (Sreedhar et al., Findings 2024)
- PDF:
- https://preview.aclanthology.org/landing_page/2024.findings-emnlp.713.pdf