ConceptAI & ML
Post-training
The second stage of training where a base model is turned into an assistant by training on conversation datasets.
Know first
Named as a prerequisite by the lectures that define this. The number is how many say so. A grey one has no page yet.
Opens up
Concepts that name this one as a prerequisite.
Lecturers
4Each in their own terms, from the easiest level.
Matt Jones
The phase of model development where system responses are trained and tweaked with the help of humans.Adam Brown
Post-training sends a pretrained neural network to finishing school to only be polite and helpful to the user rather than just predicting the next word.Percy Liang
The phase of training that occurs after pre-training, often involving fine-tuning or alignment.
Lectures that assume it
2Connections
14The lectures that define it or assume it, and the concepts next to it.
Concepts8
Show 2 moreShow fewer
Lectures6
- Deep Dive into Large Language Models: Pretraining to RLHF
- AI as Therapist? Intimacy Without Understanding
- Training Sand to Think: Artificial General Intelligence & Future of Physics
- Marin: An Open Lab For Frontier AI
- AI Seminar: Actionable Interpretability for Language Model Reasoning
- Overview of Transformers (Stanford CS25: Transformers United V6)