Saved in:
| Main Authors: | , , , , , |
|---|---|
| Format: | Preprint |
| Published: |
2025
|
| Subjects: | |
| Online Access: | https://arxiv.org/abs/2501.12698 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1866916583910146048 |
|---|---|
| author | Yoshida, Kai Mizukami, Masahiro Kawano, Seiya Kruengkrai, Canasai Sugiyama, Hiroaki Yoshino, Koichiro |
| author_facet | Yoshida, Kai Mizukami, Masahiro Kawano, Seiya Kruengkrai, Canasai Sugiyama, Hiroaki Yoshino, Koichiro |
| contents | To improve user engagement during conversations with dialogue systems, we must improve individual dialogue responses and dialogue impressions such as consistency, personality, and empathy throughout the entire dialogue. While such dialogue systems have been developing rapidly with the help of large language models (LLMs), reinforcement learning from AI feedback (RLAIF) has attracted attention to align LLM-based dialogue models for such dialogue impressions. In RLAIF, a reward model based on another LLM is used to create a training signal for an LLM-based dialogue model using zero-shot/few-shot prompting techniques. However, evaluating an entire dialogue only by prompting LLMs is challenging. In this study, the supervised fine-tuning (SFT) of LLMs prepared reward models corresponding to 12 metrics related to the impression of the entire dialogue for evaluating dialogue responses. We tuned our dialogue models using the reward model signals as feedback to improve the impression of the system. The results of automatic and human evaluations showed that tuning the dialogue model using our reward model corresponding to dialogue impression improved the evaluation of individual metrics and the naturalness of the dialogue response. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2501_12698 |
| institution | arXiv |
| publishDate | 2025 |
| record_format | arxiv |
| spellingShingle | Training Dialogue Systems by AI Feedback for Improving Overall Dialogue Impression Yoshida, Kai Mizukami, Masahiro Kawano, Seiya Kruengkrai, Canasai Sugiyama, Hiroaki Yoshino, Koichiro Computation and Language To improve user engagement during conversations with dialogue systems, we must improve individual dialogue responses and dialogue impressions such as consistency, personality, and empathy throughout the entire dialogue. While such dialogue systems have been developing rapidly with the help of large language models (LLMs), reinforcement learning from AI feedback (RLAIF) has attracted attention to align LLM-based dialogue models for such dialogue impressions. In RLAIF, a reward model based on another LLM is used to create a training signal for an LLM-based dialogue model using zero-shot/few-shot prompting techniques. However, evaluating an entire dialogue only by prompting LLMs is challenging. In this study, the supervised fine-tuning (SFT) of LLMs prepared reward models corresponding to 12 metrics related to the impression of the entire dialogue for evaluating dialogue responses. We tuned our dialogue models using the reward model signals as feedback to improve the impression of the system. The results of automatic and human evaluations showed that tuning the dialogue model using our reward model corresponding to dialogue impression improved the evaluation of individual metrics and the naturalness of the dialogue response. |
| title | Training Dialogue Systems by AI Feedback for Improving Overall Dialogue Impression |
| topic | Computation and Language |
| url | https://arxiv.org/abs/2501.12698 |