Few-Shot
النمط القليل العيّنات
إعداد تقييم يُزوَّد فيه النموذج بأمثلة محلولة قبل السؤال الفعلي لتوجيه إجاباته — يُضعف أداء نماذج الاستدلال الطويل.
An evaluation setting where the model is given solved examples before the actual question to guide its answers — degrades performance of long-reasoning models.
Also translated asالقليل العيّنات، متعدد الأمثلة
First appears in this corpus in: Language Models Are Few-Shot Learners (2020)
Appears in these papers
- BLOOM: A 176B-Parameter Open-Access Multilingual Language Model2022in the sky ✦
- Are Emergent Abilities of Large Language Models a Mirage?2023in the sky ✦
- Gemini: A Family of Highly Capable Multimodal Models2023in the sky ✦
- Language Models Are Few-Shot Learners2020in the sky ✦
- LIMA: Less Is More for Alignment2023in the sky ✦
- Self-Instruct: Aligning Language Models with Self-Generated Instructions2022in the sky ✦
- Toolformer: Language Models Can Teach Themselves to Use Tools2023in the sky ✦
- Toolformer: Language Models Can Teach Themselves to Use Tools2023in the sky ✦
- Weak-to-Strong Generalization: Eliciting Strong Capabilities With Weak Supervision2023in the sky ✦