Mikołaj Langner - Dichotomic Prompting for Efficient Multi-LabelLLM-Based Classfication
Автор: Instytut Podstaw Informatyki PAN
Загружено: 2025-12-14
Просмотров: 16
Описание:
We demonstrate that reasoning-enabled LLMs are markedly better at
tasks requiring contextual sensitivity, such as offensive-language
annotation. When prompted to adopt a specific role, reasoning models
maintain that role more consistently and make more accurate, fine-grained
judgments than their non-reasoning counterparts. Viewed together, these
findings highlight a unifying principle: LLMs become both more efficient and
more context-aware when their decision process is made more structured,
whether through task decomposition or through explicit reasoning.
Повторяем попытку...
Доступные форматы для скачивания:
Скачать видео
-
Информация по загрузке: