X · @teortaxesTex
· X / Twitter
> We did eval for Kimi K2.5 … Our conclusion for cyber: "it exhibits strong cybersecurity knowledge, competitive performance on structured offensive …
> We did eval for Kimi K2.5 … Our conclusion for cyber: "it exhibits strong cybersecurity knowledge, competitive performance on structured offensive tasks, and demonstrated capacity for semi-autonomous offensive execution…Bullish for K3! Let's see if they had sandbaggedYong Zheng-Xin: Given that the Kimi K3 will be rel