Table 8.

Similarity of representations between different layers of pairs of networks trained from random initialization (R.I.) and via transfer learning (T.L.).

	encoder-1	encoder-2	encoder-3	decoder-1	decoder-2	decoder-3
RSIM(R.I., R.I.)	0.247 ± 0.045	0.358 ± 0.034	0.405 ± 0.018	0.463 ± 0.046	0.540 ± 0.064	0.398 ± 0.041
RSIM(T.L., T.L.)	0.243 ± 0.010	0.370 ± 0.009	0.433 ± 0.005	0.510 ± 0.018	0.492 ± 0.047	0.333 ± 0.050
RSIM(R.I., T.L.)	0.287 ± 0.019	0.386 ± 0.014	0.460 ± 0.021	0.597 ± 0.024	0.636 ± 0.046	0.534 ± 0.057