Scale, alignment and safety

Scaling laws, reinforcement learning from human feedback, and the safety agenda.

Solid lines are confirmed citations (OpenAlex or the citing paper's own reference list); dotted lines are influences stated in textbooks without a confirmed citation. Vertical position is the theme group.

Year	First author	Title	Era
2013	Szegedy et al.	Intriguing properties of neural networks	Ch. 5 Deep learning 2012–17
2016	Amodei et al.	Concrete Problems in AI Safety	Ch. 5 Deep learning 2012–17
2017	Christiano et al.	Deep reinforcement learning from human preferences	Ch. 5 Deep learning 2012–17
2017	Hestness et al.	Deep Learning Scaling is Predictable, Empirically	Ch. 6 Pretraining and generation 2017–22
2020	Kaplan et al.	Scaling Laws for Neural Language Models	Ch. 6 Pretraining and generation 2017–22
2020	Stiennon et al.	Learning to summarize from human feedback	Ch. 6 Pretraining and generation 2017–22
2021	Fedus et al.	Switch Transformers: Scaling to Trillion Parameter Models with Simple and Efficient Sparsity	Ch. 6 Pretraining and generation 2017–22
2021	Bender et al.	On the Dangers of Stochastic Parrots: Can Language Models Be Too Big?	Ch. 6 Pretraining and generation 2017–22
2022	Hoffmann et al.	Training Compute-Optimal Large Language Models	Ch. 6 Pretraining and generation 2017–22
2022	Ouyang et al.	Training language models to follow instructions with human feedback	Ch. 6 Pretraining and generation 2017–22
2022	Chowdhery et al.	PaLM: Scaling Language Modeling with Pathways	Ch. 6 Pretraining and generation 2017–22
2022	Wei et al.	Emergent Abilities of Large Language Models	Ch. 6 Pretraining and generation 2017–22