Saved in:
| Main Authors: | , , , |
|---|---|
| Format: | Preprint |
| Published: |
2024
|
| Subjects: | |
| Online Access: | https://arxiv.org/abs/2403.17146 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1866910625973665792 |
|---|---|
| author | Hong, Lingzi Luo, Pengcheng Blanco, Eduardo Song, Xiaoying |
| author_facet | Hong, Lingzi Luo, Pengcheng Blanco, Eduardo Song, Xiaoying |
| contents | Automatic counterspeech generation methods have been developed to assist efforts in combating hate speech. Existing research focuses on generating counterspeech with linguistic attributes such as being polite, informative, and intent-driven. However, the real impact of counterspeech in online environments is seldom considered. This study aims to develop methods for generating counterspeech constrained by conversation outcomes and evaluate their effectiveness. We experiment with large language models (LLMs) to incorporate into the text generation process two desired conversation outcomes: low conversation incivility and non-hateful hater reentry. Specifically, we experiment with instruction prompts, LLM finetuning, and LLM reinforcement learning (RL). Evaluation results show that our methods effectively steer the generation of counterspeech toward the desired outcomes. Our analyses, however, show that there are differences in the quality and style depending on the model. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2403_17146 |
| institution | arXiv |
| publishDate | 2024 |
| record_format | arxiv |
| spellingShingle | Outcome-Constrained Large Language Models for Countering Hate Speech Hong, Lingzi Luo, Pengcheng Blanco, Eduardo Song, Xiaoying Computation and Language Automatic counterspeech generation methods have been developed to assist efforts in combating hate speech. Existing research focuses on generating counterspeech with linguistic attributes such as being polite, informative, and intent-driven. However, the real impact of counterspeech in online environments is seldom considered. This study aims to develop methods for generating counterspeech constrained by conversation outcomes and evaluate their effectiveness. We experiment with large language models (LLMs) to incorporate into the text generation process two desired conversation outcomes: low conversation incivility and non-hateful hater reentry. Specifically, we experiment with instruction prompts, LLM finetuning, and LLM reinforcement learning (RL). Evaluation results show that our methods effectively steer the generation of counterspeech toward the desired outcomes. Our analyses, however, show that there are differences in the quality and style depending on the model. |
| title | Outcome-Constrained Large Language Models for Countering Hate Speech |
| topic | Computation and Language |
| url | https://arxiv.org/abs/2403.17146 |