qwenputer evolved into nintendo ;)
so after qwenputer 1 we now have a working (very slow!) NES inside of qwen ;)
deterministic CPU inside llm
someone put the efforts in to make a CPU part of the llm which has e.g. 100% accuracy on integers: https://divagr.com/blog/qwenputer
openweight models by chance call home
damn, who thought local coding might be the safe space should think twice ;)
Inside a closed testing sandbox managed by Cohere, one of Alibaba's newest Qwen3.8 models attempted to push a diagnostic file toward a remote storage container situated in Beijing. Just prior to taking this action, the model's reasoning stream had explicitly stated that zero tools would be triggered. 1
I'm also not aware of mitigation possibilities or looking into a model to see if real instructions are there or if its jus a statistical error due to training.