Interconnects (Nathan Lambert)
· Newsletters
Latest open artifacts (#23): Laguna S2.1, Inkling, & Kimi K3 show the utility of open models on the Pareto frontier
Capacity to train strong models is proliferating.