Emergent Ability
قدرة ناشئة
قدرة لا تظهر في النماذج الصغيرة ولا يمكن التنبؤ بها من أدائها، بل تبرز فجأة عند بلوغ النموذج حجماً معيناً — كما يحدث مع الاستدلال بسلسلة التفكير عند تجاوز 100 مليار معامل.
A capability absent in small models and unpredictable from their performance, which appears suddenly once a model reaches a certain scale — as happens with chain-of-thought reasoning beyond ~100B parameters.
Also translated asخاصية ناشئة، سلوك طارئ مرتبط بالحجم
First appears in this corpus in: Improving Language Understanding by Generative Pre-Training (2018)
Appears in these papers
- Chain-of-Thought Prompting Elicits Reasoning in Large Language Models2022in the sky ✦
- Chain-of-Thought Prompting Elicits Reasoning in Large Language Models2022in the sky ✦
- Emergent Abilities of Large Language Models2022in the sky ✦
- Improving Language Understanding by Generative Pre-Training2018in the sky ✦
- Language Models Are Unsupervised Multitask Learners2019in the sky ✦
- PaLM: Scaling Language Modeling with Pathways2022in the sky ✦
- PaLM: Scaling Language Modeling with Pathways2022in the sky ✦