Revision History
Momentum · 11 revisions
Sizes are character counts of the article source. The signed number is the change from the previous revision.
Recent edit summaries
Detailed summaries recorded by editors. Generic maintenance summaries are omitted here; every recorded revision remains below. Dates describe the edit, not necessarily the event it covers.
- Correction: Corrected the TensorFlow/Keras SGD momentum formula (it folds the learning rate into the velocity), removed PaLM from the AdamW LLM-pretraining row (PaLM used Adafactor) and fixed the GPT-3 learning-rate range, and corrected the attribution of the sparse initialization scheme to Martens (2010) rather than Sutskever et al (plus 8 minor fixes)
Version 11 · Nov 2, 2026, 08:18 AM
- Update: added a video.
Version 10 · Nov 1, 2026, 08:16 AM
- Improve internal links (2026-10-13)
Version 8 · Oct 13, 2026, 10:04 AM
- Added 6 contextual internal links
Version 7 · Jul 23, 2026, 03:02 PM
- Render formulas with LaTeX math notation
Version 6 · Jul 11, 2026, 06:09 PM