Revision History
Adversarial attack · 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 a composite Goodfellow quote, a fabricated 90% Face++ success rate, an unsupported C&W result attributed to a brain-tumor study, wrong ImageNet RobustBench figures and a CIFAR-100 result misattributed to ImageNet, overstated Best-of-N jailbreak success rates, wrong TextFooler numbers, a misdescribed many-shot jailbreak mitigation, and the 2026 (not 2025) date of Anthropic's next-generation Constitutional Classifiers (plus 20 minor fixes)
Version 11 · Nov 2, 2026, 12:49 PM
- Update: added a video.
Version 10 · Nov 1, 2026, 08:19 AM
- Improve internal links (2026-10-13)
Version 9 · Oct 13, 2026, 10:07 AM
- Added 6 contextual internal links
Version 8 · Jul 23, 2026, 03:03 PM
- Render formulas with LaTeX math notation
Version 7 · Jul 11, 2026, 07:13 PM