AI roadmap

AI Papers Reading Roadmap

A focused, no-login list of foundational and frontier papers. Open any paper directly in a new tab. This page is static and does not collect data.

18 papers 7 phases 2014-2025

Phase 1 - Foundations

01 Sequence to Sequence Learning with Neural Networks2014 Sutskever et al. - Encoder-decoder backbone used before Transformers.
02 Neural Machine Translation by Jointly Learning to Align and Translate2014 Bahdanau et al. - Introduced attention, a key prerequisite for Transformers.

Phase 2 - The Transformer Era

Phase 3 - Scaling and Few-Shot Learning

Phase 4 - Alignment and RLHF

Phase 5 - Efficient and Open Models

Phase 6 - Chain-of-Thought and Reasoning

Phase 7 - Frontier (2024-2025)