Axel Möllerberg on Mythryl: “”We answer this by literally building a computer inside a…”

”We answer this by literally building a computer inside a transformer. We turn arbitrary C code into tokens that the model itself can execute reliably for millions of steps in seconds.” You did what now?

percepta.aiCan LLMs Be Computers? | PerceptaWe build a computer inside a transformer — executing arbitrary C programs for millions of steps with exponentially faster inference via 2D attention heads.

13/3/26, 6:41 pm0 reposts0 likes

Post