INDEX
Explanations
small, tiny, clustered items
New Auto-Interp
Negative Logits
oversized
0.57
oversize
0.54
grandi
0.47
ใหญ่
0.44
大き
0.44
enlarge
0.41
enlarged
0.41
大型
0.41
LARGE
0.40
enlarging
0.40
POSITIVE LOGITS
tight
0.48
تالي
0.43
istern
0.41
صغير
0.41
малень
0.41
Small
0.40
tight
0.40
compaction
0.40
dicht
0.40
Tight
0.39
Activations Density 0.001%