按年被提到
20232023 · 1 次2026
大模型与语言模型›大模型预训练与缩放·arXiv:2305.15425 · cs.CL

Language Model Tokenizers Introduce Unfairness Between Languages

Aleksandar Petrov, Emanuele La Malfa, Philip H. S. Torr, Adel Bibi · 2023
开始阅读 ↗PDFalphaXiv 讨论
被提及 · 2 人首次 2023-08-03 · 最近 2026-09-10
2不同的人
2活跃月份
2讨论串
78被引(OpenAlex)
arXiv 摘要(候选层不写导读)

Recent language models have shown impressive multilingual performance, even when not explicitly trained for it. Despite this, there are concerns about the quality of their outputs across different languages. In this paper, we show how disparity in the treatment of different languages arises at the tokenization stage, well before a model is even invoked. The same text translated into different languages can have drastically different tokenization lengths, with differences up to 15 times in some cases. These disparities persist even for tokenizers that are intentionally trained for multilingual support. Character-level and byte-level models also exhibit over 4 times the difference in the encoding length for some language pairs. This induces unfair treatment for some language communities in regard to the cost of accessing commercial language services, the processing time and latency, as well as the amount of content that can be provided as context to the models. Therefore, we make the case that we should train future language models using multilingually fair subword tokenizers.

候选层:至少 2 个不同的人、在 2 个不同月份提到过,还没到主书架的 3 人 3 个月。没有做立场标注,所以只写“被提及”。

HN 书评

推荐在前,同类按时间 · 每篇最多保留 20 条,统计用全部提及
评论fnordpiglet · 2023-08

“Given the performance of tokenizers by language the answer must be “worse,” there are a variety of sources for this.. they typically focus on cost, but it’s also borne out in quality. https://denyslinkov.medium.com/why-is-gpt-3-15-77x-more-expensive-for-certain-languages-2b19a4adc4bc https://arxiv.org/pdf/2305.13707 https://arxiv.org/pdf/2305.15425 A…”

被提交为 story · 1 次