SynthDetoxM: Modern LLMs are Few-Shot Parallel Detoxification Data Annotators

🤗 Upvotes: 71 | cs.CL

Authors:
Daniil Moskovskiy, Nikita Sushko, Sergey Pletenev, Elena Tutubalina, Alexander Panchenko

Title:
SynthDetoxM: Modern LLMs are Few-Shot Parallel Detoxification Data Annotators

Arxiv:
http://arxiv.org/abs/2502.06394v1

Abstract:
Existing approaches to multilingual text detoxification are hampered by the scarcity of parallel multilingual datasets. In this work, we introduce a pipeline for the generation of multilingual parallel detoxification data. We also introduce SynthDetoxM, a manually collected and synthetically generated multilingual parallel text detoxification dataset comprising 16,000 high-quality detoxification sentence pairs across German, French, Spanish and Russian. The data was sourced from different toxicity evaluation datasets and then rewritten with nine modern open-source LLMs in few-shot setting. Our experiments demonstrate that models trained on the produced synthetic datasets have superior performance to those trained on the human-annotated MultiParaDetox dataset even in data limited setting. Models trained on SynthDetoxM outperform all evaluated LLMs in few-shot setting. We release our dataset and code to help further research in multilingual text detoxification.