Introducing SubQ - a major breakthrough in LLM intelligence.
It is the first model built on a fully sub-quadratic sparse-attention architecture (SSA),
And the first frontier model with a 12 million token context window which is:
- 52x faster than FlashAttention at 1MM tokens
- Less than 5% the cost of Opus
Transformer-based LLMs waste compute by processing every possible relationship between words (standard attention).
Only a small fraction actually matter.
@subquadratic finds and focuses only on the ones that do.
That's nearly 1,000x less compute and a new way for LLMs to scale.
@BitcoinMagazine@Duelbits Sorry but this is not good that he is seemingly the only buyer and concentrating the market. @saylor needs to find ways for others to buy bitcoin not buy it all himself and risk fucking it all up…
🚨“WE GOT HIM! My fellow Americans, over the past several hours, the United States Military pulled off one of the most daring Search and Rescue Operations in U.S. History, for one of our incredible Crew Office Members, who also happens to be a highly respected Colonel, and who I am thrilled to let you know is SAFE and SOUND!” - President Donald J. Trump 🇺🇸
@LynAldenContact I feel sad that @PeterMcCormack handed over the ranks of WBD podcast (no offense to @_DannyKnowles). I always looked forward to it. Also @nlw handing over The Breakdown broke my heart. Everyone I relied on through the bear market to calm me left the space to next shiny new thing
Why should the red states pay for the corruption and mismanagement of California? Let the bond market dry up and retirement benefits go into default. Let’s flip CA red for 200 years.
“In the beginning of a change the patriot is a scarce man, and brave, and hated and scorned. When his cause succeeds, the timid join him, for then it costs nothing to be a patriot.” ~ Big T