INDEX
Explanations
phrases related to time, particularly nighttime hours
New Auto-Interp
Neuron Alignment
Index
Value
% of L₁
897
+0.13
0.4%
1013
+0.10
0.3%
946
+0.10
0.3%
Correlated Neurons
Index
P. Corr.
Cos Sim.
1062
+0.13
0.04
1443
+0.10
0.03
655
+0.10
0.04
Negative Logits
affez
-0.61
znál
-0.61
PÉ
-0.49
cammin
-0.48
espri
-0.48
soggior
-0.47
redus
-0.47
gius
-0.46
kado
-0.46
Warto
-0.45
POSITIVE LOGITS
night
0.64
midnight
0.64
midnight
0.63
overnight
0.63
night
0.61
nights
0.60
nocturnal
0.60
Overnight
0.58
nighttime
0.58
Midnight
0.57
Activations Density 0.132%