Unverified50% confidenceFactExact time
主流模型如 GPT-4、Claude 在训练阶段被强化了对系统提示词的优先遵从性，系统提示词的约束效力通常强于用户指令
1
Sources
50%
Confidence
Long-term
Relevance
7/9/2026
First Seen
Sources
Related Claims
UnverifiedClaude tends to follow detailed system prompts, while the GPT series responds better to structured instructions.81% similarUnverifiedMany open-source models use high-quality outputs generated by commercial models such as GPT-4 and Claude as training data, which is the knowledge distillation training paradigm.80% similarUnverifiedMainstream large language models (including GPT-4, DeepSeek, Claude, etc.) are all trained on static corpora and have a definitive knowledge cutoff date.79% similarUnverifiedGPT-4、Claude等通用大模型在预训练阶段学习的是互联网上的海量公开文本，倾向于给出通用的回答79% similarVerifiedModels such as GPT-4 and Claude 3 already have sufficient capabilities in reasoning, planning, and instruction-following to support reliable Agent applications.77% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/401455API
curl https://kongchang.com/api/v1/knowledge/claims/401455MCP
get_claim(id=401455)