Understanding Transformers via N-gram Statistics

71 pona-a 1 5/17/2025, 7:56:00 PM arxiv.org ↗

Comments (1)

justanotherjoe · 21m ago
Sounds regressive and feeds into the weird unintellectual narrative that llm is just like ngram models (lol, lmao even)

Thr author submitted like 10 papers this May alone. Is that weird?