Why feed 1M tokens when ~250k visual tokens do? 🚀👀 Concurrent to DeepSeek-OCR, today we’re releasing Glyph, a visual-text compression paradigm that turns long text into images and lets a VLM read them. Paper: https://t.co/dvYaKjWoXW @karpathy may be you will be also https://t.co/mEY4BWCJMJ
— Xiao Liu (Shaw) (@ShawLiu12) Oct 21, 2025
from Twitter https://twitter.com/ShawLiu12
October 21, 2025 at 04:06AM
via IFTTT
No comments:
Post a Comment