Quadrillion-parameter transformers and how to train them

Re:Quadrillion-parameter transformers and how to train them by smokel on Sunday November 08, 2026 @10:00AM (#50413881)

And several thousand quadrillon grains of sand.

[ Parent | on HN ]

Re:Quadrillion-parameter transformers and how to train them by intrasight on Sunday November 08, 2026 @10:10AM (#50413960)

There are more ways to shuffle a deck of cards than there are atoms in the visible universe.

[ Parent | on HN ]

Re:Quadrillion-parameter transformers and how to train them by smokel on Sunday November 08, 2026 @10:20AM (#50414025)

A typical grain of sand contains in the order of hundreds of quadrillions of atoms. A deck of cards can be encoded in 226 bits per deck, so one could encode in the order of a quadrillion decks of cards in a single grain of sand!

[ Parent | on HN ]

Re:Quadrillion-parameter transformers and how to train them by mstaoru on Sunday November 08, 2026 @10:58AM (#50414338)

What if it's a MtG deck?

[ Parent | on HN ]

Re:Quadrillion-parameter transformers and how to train them by amelius on Sunday November 08, 2026 @10:56AM (#50414324)

There are more ways to configure a single atom than there are card shuffles of a single deck.

[ Parent | on HN ]

Re:Quadrillion-parameter transformers and how to train them by aqme28 on Sunday November 08, 2026 @02:08PM (#50416047)

what do you mean configure?

[ Parent | on HN ]

Re:Quadrillion-parameter transformers and how to train them by hansvm on Sunday November 08, 2026 @02:13PM (#50416091)

I could have sworn there were infinite (but as energies increase they become increasingly easy to ionize).

[ Parent | on HN ]

Re:Quadrillion-parameter transformers and how to train them by aqme28 on Sunday November 08, 2026 @03:48PM (#50417058)

Maybe for some theoretical ideal but definitely not in practice.

[ Parent | on HN ]