They did reinforcement learning on Kimi K3, probably specifically targeting the benchmarks. Eh, it isn't news. I mean, it's just an echo bit of news to the gre

They did reinforcement learning on Kimi K3, probably specifically targeting the benchmarks. Eh, it isn't news. I mean, it's just an echo bit of news to the great K3 release.