LLMilestonesA milestone timeline of AI and large language models

Act III · The ChatGPT moment

PaperCMU / Princeton

Mamba: selective state-space models

A sequence model whose cost grows linearly with sequence length, matching Transformers of the same size.

Why it matters

Revived research into alternatives to attention; hybrids of linear layers and attention later went into production models.

Led to

Qwen3.5

Source: arxiv.org ↗

See it on the 3D timeline →