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

Unverified95% confidenceFactTime unknown

大模型核心原理包括Transformer架构、预训练、SFT监督微调、RLHF人类反馈强化学习等关键概念

1

Sources

95%

Confidence

Long-term

Relevance

6/1/2026

First Seen

Sources

2026年AI学习路线图：3条落地路径+四步行动清单（附完整技术体系）

bilibiliAi大模型精讲课堂

Related Entities

Transformerconcept SFTconcept RLHFconcept

Related Claims

Verified指令微调技术由InstructGPT（2022）系统化提出，结合人类反馈强化学习（RLHF）使模型更符合人类意图78% similar Unverified领域微调通常先通过监督微调（SFT）注入领域知识，再经过基于人类反馈的强化学习（RLHF）对齐用户偏好73% similar Unverified大模型（LLM）是基于Transformer架构、经过海量数据预训练的深度神经网络模型73% similar Unverified现代LLM通过指令微调（Instruction Tuning）和基于人类反馈的强化学习（RLHF）获得了规划、推理和工具使用能力73% similar Unverified学习Transformer应遵循'先建立问题意识，再引入解决方案'的四阶段路径：序列建模动机、RNN/LSTM及其局限、注意力机制、Transformer完整架构73% similar

Cite This Claim

Stable URI

https://kongchang.com/claim/20726

API

curl https://kongchang.com/api/v1/knowledge/claims/20726

MCP

get_claim(id=20726)