Hi, there doesn't seem to be code corresponding to the memory palace task, or Fig. 7 in the paper? Could you kindly share the code you used to train the model for the task, and for generating the plots in Fig 7? Curious to know how the sensory-HPC weights are learned for the mem palace task, as we couldn't find specific info in the paper. Thank you so much in advance!