Fig. 2.

fMRI decoding results in bilateral EC and subiculum. a Using a linear SVC with L2 regularisation, in bilateral posterior EC group decoding accuracy for allocentric boundary direction was significantly above chance, as determined by a boot-strap permutation test (10,000 samples). It was not possible, however, to decode allocentric goal direction. b In anterior EC, we were able to decode both allocentric boundary and goal direction. c The posterior subiculum contained similar information to the posterior EC, in that we could decode allocentric boundary, but not goal, direction. d Finally, in anterior subiculum it was possible to decode allocentric goal, but not boundary, direction. All p values were determined via non-parametric Monte Carlo significance tests. Source data are provided as a Source Data file