教程中心
AI Agent 从入门到实战：概念理解、MCP 使用、平台实操、工作流自动化
1252
教程总数
234
入门教程
42
实操教程
按主题浏览
Diffusion Models Explained: From DDPM to Stable Diffusion and FLUX
Technical walkthrough of denoising diffusion, latent spaces, and conditioning mechanisms
扩散模型技术详解（2026）：从 DDPM 前向/反向过程数学、UNet 噪声预测，到 Latent Diffusion（Stable Diffusion 的 64 倍效率技巧）、CFG 引导公式，再到 DiT 与 rectified flow（SD3/FLUX）。附生态工具与理论的对照表。
RLHF vs DPO: Training LLMs from Human Feedback - Technical Guide 2025
Reinforcement Learning from Human Feedback, Direct Preference Optimization, and alternatives
RLHF vs DPO 偏好学习指南（2026）：把基座模型对齐成有用/无害/诚实的助手。RLHF 三阶段(SFT+奖励模型+PPO)复杂但强；DPO 用单一偏好损失省去奖励模型与 RL、更稳更简。含选型表与 IPO/KTO 等变体。
Causal Inference for ML Engineers: Treatment Effects, Uplift Modeling, and A/B Testing
DoWhy, CausalML, and production causal modeling for data-driven decisions
ML 工程师的因果推断（2026）：用潜在结果框架回答"改变 X 会不会导致 Y"。涵盖 A/B、倾向得分匹配、工具变量、双重差分、Double ML 与 uplift 建模，及 DoWhy/CausalML/EconML 库。
AI Model Interpretability: SHAP, LIME, and Integrated Gradients for XAI
Explaining black-box ML models for compliance, debugging, and stakeholder communication
Master explainable AI techniques including SHAP values, LIME, integrated gradients, and attention visualization to interpret machine learning models for debugging, compliance, and stakeholder communication.
Reinforcement Learning for Real-World Applications: Beyond Game AI
Production RL for robotics, resource optimization, and recommendation systems
Learn practical reinforcement learning applications beyond games including supply chain optimization, cloud resource management, recommendation systems, and robotics control with modern RL libraries.
Deep Learning for Tabular Data: When Neural Nets Beat Gradient Boosting
TabNet, FT-Transformer, and AutoML approaches for structured data problems
Explore when and how deep learning approaches (TabNet, FT-Transformer, SAINT) outperform gradient boosting on tabular data, with practical implementation and hyperparameter guidance.
Federated Learning in Practice: Training AI Models Without Centralizing Data
Flower framework, differential privacy, and production FL for mobile and edge devices
Practical guide to federated learning using the Flower framework, covering federation strategies, differential privacy, communication efficiency, and real-world deployment for healthcare and fintech.
Advanced Time Series Forecasting with AI: N-BEATS, PatchTST, and Foundation Models
From classical ARIMA to neural and foundation model approaches for production forecasting
Comprehensive guide to advanced time series forecasting using neural architectures including N-BEATS, PatchTST, Chronos, and TimeGPT, with practical implementation and model selection guidance.
Knowledge Distillation: Train Small, Fast AI Models from Large Teacher Models
Task-specific distillation, intermediate layer matching, and deployment tradeoffs
Learn knowledge distillation techniques to create small, fast student models that mimic large teacher model performance, covering task distillation, feature-level distillation, and production deployment.
Foundation Models for Robotics: RT-2, OpenVLA, and Physical Intelligence
How vision-language-action models are enabling general-purpose robot control
Explore how foundation models are transforming robotics through vision-language-action (VLA) models like RT-2 and OpenVLA, enabling robots to follow natural language instructions and generalize to new tasks.
AI Model Merging: SLERP, TIES, DARE, and Model Soup Techniques
Combine multiple fine-tuned models without additional training to create superior models
Explore model merging techniques that combine weights from multiple fine-tuned models to create superior models without additional training, including SLERP, TIES-Merging, DARE, and evolutionary approaches.
Graph Neural Networks in Production: Applications, Architectures, and Best Practices
GCN, GAT, GraphSAGE for fraud detection, recommendation, and molecular design
Learn practical applications of Graph Neural Networks including fraud detection in financial transactions, molecule property prediction, knowledge graph completion, and large-scale recommendation systems.
Transformer Architecture Deep Dive: Attention Mechanisms and Modern Variants
From vanilla attention to Flash Attention, Grouped Query Attention, and Mamba
Comprehensive technical deep dive into transformer architecture including self-attention, multi-head attention, positional encoding, and modern efficiency improvements used in GPT-4 and Llama.