"Only large models have emergent abilities" This mystic mantra is challenged again Broad 🧵 on the debate + new findings: 165M parameter models trained on simple English are better zero shot…

Posters: LChoshen. LChoshen

2 likes recorded by the source corpus.