INDEX
Explanations
personal anecdotes and experiences
The neuron fires on positive, forward‐looking expressions of personal enthusiasm or intent—words like “excited,” “thrilled,” and “planning.”
New Auto-Interp
Negative Logits
distribution
-0.07
JPEG
-0.07
if
-0.06
سان
-0.06
"}, ↵
-0.06
devised
-0.06
Size
-0.06
}(
-0.06
distance
-0.06
.E
-0.06
POSITIVE LOGITS
ộc
0.07
ourney
0.06
roys
0.06
("@0.06
.sep
0.06
_MISSING
0.06
-agent
0.06
-http
0.06
gpointer
0.06
gouver
0.06
Activations Density 0.101%