Revision History
Reasoning models · 7 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.
- Bring the reasoning models survey current through July 2026: add a Control surfaces section (token budgets, effort levels, adaptive thinking, what providers return), extend training methods with RLVR at scale and the elicit-versus-teach debate, rebuild the representative systems table from 5 to 21 rows, and add benchmark saturation, contamination, overthinking and the illusion-of-thinking debate with its rebuttals
Version 7 · Aug 1, 2026, 11:16 AM
- Correction: reverified Reasoning models with bounded terminology, historical precursors, disclosed training and inference methods, protocol-aware evaluation, and qualified limitations; replaced unsupported chronology and cross-model rankings with 13 primary and academic sources.
Version 6 · Jul 29, 2026, 07:50 AM
- Added 6 contextual internal links
Version 5 · Jul 23, 2026, 03:30 PM