A linear-attention naming joke, from Linformer to Qwen3.8
A user jokes that models using linear attention once had names like Linformer, but now they’re called Qwen3.8.
TLDR
A user quips that models using linear attention used to have names “something goofy like Linformer.” The punchline: “Now they’re just called Qwen3.8.”
Combined views
—
1 Source, first seen ago
likes
