课程

Explore how LLMs work, from transformer architecture to emergent capabilities and real-world limitations.

Master advanced prompting techniques from zero-shot to tree-of-thought, and learn to build safe, reusable prompt templates.

Understand the three major deep learning architectures and why transformers came to dominate modern AI.

Learn when and how to fine-tune models or use retrieval-augmented generation to build domain-specific AI applications.

Discover how AI agents observe, reason, and act - and why multi-agent collaboration is shaping the future of artificial intelligence.

A deep dive into self-attention, multi-head attention, positional encoding, and the Transformer architecture that powers every modern large language model.

Explore the empirical scaling laws that govern model performance, compute-optimal training strategies, and the distributed systems engineering behind training frontier models.

Understand how raw language models are transformed into helpful, harmless assistants through supervised fine-tuning, reward modelling, and reinforcement learning from human feedback.

Learn how AI agents observe, reason, and act autonomously using tool use, memory, planning strategies, and multi-agent architectures for complex real-world tasks.

Explore the attack surface of modern AI systems - from jailbreaks and prompt injection to adversarial examples and data poisoning - and learn the defence strategies that protect them.