State-space models can learn in-context by gradient descent

Re-assign community