The Decade of Attention
How Transformers Rebuilt Machine Learning, 2017–2027
By Dr. Priya Iyengar
No part of this publication may be reproduced, stored in a retrieval system, or transmitted in any form or by any means — electronic, mechanical, photocopying, recording, or otherwise — without the prior written permission of the publisher, except for brief quotations in reviews …
§ Contents
- 01Before Attention: The Sequence Problem
- 02The Paper
- 03The Pretraining Era: BERT, GPT-1, T5
- 04Scaling Up: GPT-2 and GPT-3
- 05The Compute Wall and Chinchilla's Lessons
- 06Alignment and ChatGPT
- 07The Open-Weights Awakening
- 08Architectural Variants: MoE and the State-Space Detour
- 09Multimodality and the Universal Encoder
- 10The Inference Frontier
- 11Reasoning, Tool Use, and Agents
- 12After Attention
§ Editions
Digital · PDF
Free · No paywall
- Sold outAudioISBN 978-1-7398-2242-8
Paperback
Coming soon · Ships worldwide