Nathan Lambert 评 Kimi K3.1 Pro 基准
TLDR on the kimi k3.1 pro benchmarks is that it looks like they figured out their RL, where the initial release was a very impressive base model with lighter po
AI 导读
Nathan Lambert 认为，Kimi K3.1 Pro 的基准表现说明其团队已经解决了 RL 环节：初版是一个很强的基座模型，但后训练较轻。他还表示，中国在能力提升上正紧跟节奏。
正文
TLDR on the kimi k3.1 pro benchmarks is that it looks like they figured out their RL, where the initial release was a very impressive base model with lighter post-training. China is very much keeping pace with capability gains.