Forgetting may be the secret to better AI language learning

Дата публикации: 23-06-2026 18:40:18

Giving AI a human-like memory limitation may actually help it learn language better. In their new proof-of-principle study, Abishek Thamma (University of Amsterdam) and Micha Heilbron (Max Planck Institute for Psycholinguistics) show that small language models equipped with a transient memory learn grammar more efficiently when trained on child-scale amounts of language input. The findings demonstrate how insights from psycholinguistics can inspire new approaches to AI learning. The findings are published in the journal Transactions of the Association for Computational Linguistics.

Классификация: Наука

Схожие новости

#	Наименование новости
1	Artificial synapse uses light-color programming for brain-like balanced learning
2	Brain-inspired AI architecture could computing faster and far less power-hungry
3	Users trust AI and human fact-checkers equally, but for different reasons
4	Why AI Citation Registries Are Emerging as a Practical Path Forward for Government Agencies
5	The consequences of relying on AI for accurate news
6	Физики из Петрозаводска нашли новый метод обучения нейросети распознаванию образов
7	Una nueva herramienta detecta cuándo los modelos de IA de lenguaje visual están "alucinando"
8	El cerebro humano arma las palabras en bloques separados como una inteligencia artificial
9	OpenAI says ChatGPT’s memory feature is getting smarter and coming to free users
10	В Югре нейросеть научили озвучивать мансийские фразы

• techxplore.com