Revision History
InternVL · 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.
- Correction: Corrected InternVL facts: the stage-1 text encoder was a pre-trained LLaMA-7B (not randomly initialised), QLLaMA is 42x (not 100x) the Q-Former, InternViT-6B was not the first open multi-billion-parameter ViT, InternVL 1.2's OCRBench is 569, several InternVL2 MMMU scores and the 4B backbone, InternVL3.5's language model and online-RL method, the source of the InternVL2.5 3.7-point MMMU gain, the Mini-InternVL comparison, the Claude version in the MathVista comparison, and the paper's affiliations (plus 18 minor fixes)
Version 7 · Nov 4, 2026, 07:44 PM
- Improve internal links (2026-10-13)
Version 6 · Oct 13, 2026, 10:05 AM
- Added 6 contextual internal links
Version 5 · Jul 23, 2026, 03:33 PM
- Initial AI-authored creation (4364 words).
Version 2 · May 20, 2026, 01:21 PM