"On January 21, San Francisco-based startup Logical Intelligence appointed LeCun to its board. Building on a theory conceived by LeCun two decades prior, the startup claims to have developed a different form of AI, better equipped to learn, reason, and self-correct. Logical Intelligence has developed what's known as an energy-based reasoning model (EBM). Whereas LLMs effectively predict the most likely next word in a sequence, EBMs absorb a set of parameters-say, the rules to sudoku-and complete a task within those confines."
"Whereas LLMs effectively predict the most likely next word in a sequence, EBMs absorb a set of parameters-say, the rules to sudoku-and complete a task within those confines. This method is supposed to eliminate mistakes and require far less compute, because there's less trial and error. The startup's debut model, Kona 1.0, can solve sudoku puzzles many times faster than the world's leading LLMs, despite the fact that it runs on just a single Nvidia H100 GPU."
Yann LeCun left Meta and has challenged the dominant belief that large language models will produce AGI. Logical Intelligence appointed LeCun to its board and claims to implement a decades-old theory in a working energy-based reasoning model (EBM). EBMs take explicit problem parameters and complete tasks within those constraints, reducing trial-and-error and mistakes while lowering compute requirements. Kona 1.0, the startup's debut model, reportedly solves sudoku puzzles many times faster than top LLMs while running on a single Nvidia H100 GPU. Target applications include error-intolerant tasks like energy-grid optimization and advanced manufacturing automation.
Read at WIRED
Unable to calculate read time
Collection
[
|
...
]