#ReadingList #CoT

CoT-Reading-List

基础论文

• Chain-of-Thought Prompting Elicits Reasoning in Large Language Models
• Large Language Models are Zero-Shot Reasoners
• Automatic Chain of Thought Prompting in Large Language Models

问题分解

• Least-to-Most Prompting Enables Complex Reasoning in Large Language Models
• Measuring and Narrowing the Compositionality Gap in Language Models

融合预测

• Self-Consistency Improves Chain of Thought Reasoning in Language Models
• Active Prompting with Chain-of-Thought for Large Language Models
• Rationale-Augmented Ensembles in Language Models

生成-校验

• STaR: Self-Taught Reasoner Bootstrapping Reasoning With Reasoning
• On the Advance of Making Language Models Better Reasoners

多语言

• Language Models are Multilingual Chain-of-Thought Reasoners

大模型背景

• PaLM: Scaling Language Modeling with Pathways
• Emergent Abilities of Large Language Models
• Language Model Cascades

Points to know

• [[Sampling Methods]]
  • Top-k Sampling
  • Nucleus Sampling
  • Beam Search
  • Temperature Sampling
• [[Decoding Methods]]
  • Greedy Decoding
  • Self-Consistency
  • [[CoT-Decoding]]
• Instruction-tuned
• System I
• System II