FundamentalsUpdated 2026-10-24
Top-p Sampling
Sample only from the smallest set of tokens whose cumulative probability reaches p.
#参数#采样#生成
In plain language
Top-p, or nucleus sampling, controls randomness alongside temperature. Smaller p means fewer candidates and more conservative output; larger p allows richer expression. Usually adjust one, not both aggressively.
Example
Use p=0.9 for factual Q&A and 0.95-1.0 for creative writing.
Recommended tools
Related terms
FundamentalsFeatured
Large Language Model
A model trained on massive text to understand and generate natural language.
3 related terms
FundamentalsFeatured
Prompt Engineering
Designing instructions, examples and constraints so a model reliably produces the desired output.
3 related terms
Fundamentals
Temperature
A parameter controlling output randomness; higher values produce more diverse results.
3 related terms
Applications
Agentic Workflow
Breaking a task into steps planned by AI, executed by tools and reviewed by humans.
3 related terms