Feed Entities Podcasts Deep Dives Developers About

[KongchangAI]

Real-time AI knowledge base. Every claim is traceable to its source, with data freshness under 24 hours.

Powered by 240+ first-hand AI sources

Knowledge

Feed
Hot Entities
Podcasts

Developers

CLI
MCP Server
REST API
About

© 2026 KongchangAI. The freshest, most accurate AI knowledge

Unverified50% confidenceFactExact time

传统控制模型多样性的方法是通过调整温度、Top-p等采样参数

1

Sources

50%

Confidence

Long-term

Relevance

7/14/2026

First Seen

Sources

口语化采样登顶ICML：提示工程配得上顶会吗？

Related Claims

Verified传统分类模型可通过温度缩放（Temperature Scaling）等后处理方法改善校准80% similar Unverified对模型能力评估必须固定推理参数（temperature、top-p等），否则同一模型在不同参数配置下可能呈现判若云泥的表现77% similar Unverified提示工程、温度参数调节与输出约束机制是应对模型表达失控问题最常用的实用手段73% similar Unverified大语言模型输出本质上是概率性的，采样温度（Temperature）、Top-P 等参数使模型每次生成内容可能不同，是软件测试的挑战72% similar Verified大语言模型在生成每个token时从概率分布中采样，temperature参数控制随机性的强弱，temperature越高输出越多样，越低则越确定71% similar

Cite This Claim

Stable URI

https://kongchang.com/claim/503933

API

curl https://kongchang.com/api/v1/knowledge/claims/503933

MCP

get_claim(id=503933)