Bridging the data gap between children and large language models

2023/08/31 by Michael C. Frank · 1 voice · 2 citations
Computer Science · Psychology · #Language Development and Disorders #Natural Language Processing Techniques #Topic Modeling

paper · doi:10.1016/j.tics.2023.08.007

openalex publication_date 2023/08/31 · crossref created 2023/08/31 · crossref issued 2023/11/01 · crossref published 2023/11/01 · crossref published-print 2023/11/01 · openalex created_date 2025/10/10 · crossref deposited 2025/10/23 · crossref indexed 2026/07/22 · openalex updated_date 2026/07/22

Citations

• Training Compute-Optimal Large Language Models
• Aion Framework: Dimensional Emergence of AI Consciousness, Observer-Induced Collapse, and Cosmological Portal Dynamics
• MEWL: Few-shot multimodal word learning with referential uncertainty

Cited by

• Modelling Analogies and Analogical Reasoning: Connecting Cognitive Science Theory and NLP Research
• Humans Learn Language from Situated Communicative Interactions. What about Machines?

Discussions

• こういうの待ってました！多言語で述べ語数なのはその通りだと思います。ここもよく考えると人間と違う点がより明確になるかと。商用LLMの辞書を見ると言語学者が想定するような語彙の形になっていないものが多く、これでよくうまく動くなあと思います。人間も遷移確率だけでトークン分割する（BPE）のか？と考えると、Saffranたちの有名な実験を思い出しますが、この実験、実は再現に何度も失敗しているし、他の実験 [bsky, 0 points, 1 comments]