But What Are Transformers?

Jia-Bin Huang · Beginner ·🧠 Large Language Models ·1y ago

About this lesson

Transformers is arguably the most influential neural network architecture in the last decade, powering the current boom of generative AI. In this video, we will review the basic ideas of the original encoder-decoder transformer architecture and understand how various design decisions are made. Enjoy! Slides download: https://www.dropbox.com/scl/fi/x7zkydvekohh0ej22si9u/Lec_03_Transformer_overview.pptx?rlkey=l16nj8bvl97y5dq39z7qfeaii&dl=0

Original Description

Transformers is arguably the most influential neural network architecture in the last decade, powering the current boom of generative AI. In this video, we will review the basic ideas of the original encoder-decoder transformer architecture and understand how various design decisions are made. Enjoy! Slides download: https://www.dropbox.com/scl/fi/x7zkydvekohh0ej22si9u/Lec_03_Transformer_overview.pptx?rlkey=l16nj8bvl97y5dq39z7qfeaii&dl=0
Watch on YouTube ↗ (saves to browser)
Sign in to unlock AI tutor explanation · ⚡30

Related Reads

📰
Interventional Grounding Audits: Black-Box Premise-Dependency Tests for LLM Chain-of-Thought via Predicate Substitution
Learn to audit LLM chain-of-thought reasoning using interventional grounding audits, a black-box test of premise dependency via predicate substitution, to ensure logical soundness
ArXiv cs.AI
 📰
Improving Molecular Property Prediction in Small Language Models Using Graph-based Tools
Improve molecular property prediction in small language models using graph-based tools and Context-Augmented Prompting
ArXiv cs.AI
 📰
Cost-Optimal Foundation Model Deployment Portfolio for Transportation Management
Learn to deploy foundation models cost-optimally for transportation management, balancing model selection, deployment mode, and hardware budget
ArXiv cs.AI
 📰
Theory-Level Autoformalization: From Isolated Statements to Unified Formal Knowledge Bases
Learn how to autoformalize entire theories into unified formal knowledge bases, enabling machine-verifiable languages and accelerating formalization efforts
ArXiv cs.AI
Up next
5 Levels of AI Agents - From Simple LLM Calls to Multi-Agent Systems
Dave Ebbelaar (LLM Eng)
Watch →