Revision History
Dropout · 10 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: the original Transformer paper describes dropout in two places (sub-layer outputs and embedding sums), not three; attention dropout came from later implementations
Version 10 · Nov 4, 2026, 03:06 PM
- Correction: Removed an unsupported GPT-3 dropout figure and a false Chinchilla no-dropout claim, corrected an overstated Reuters result, and fixed the misattribution of the recurrent-network dropout problem to the original dropout paper (plus 16 minor fixes)
Version 9 · Nov 2, 2026, 10:34 AM
- Update: added a video.
Version 8 · Nov 1, 2026, 08:18 AM
- Improve internal links (2026-10-13)
Version 6 · Oct 13, 2026, 10:08 AM
- Added 6 contextual internal links
Version 5 · Jul 23, 2026, 03:04 PM