INDEX

Explanations

various kinds of

The neuron specifically detects the appearance of the word “various.”

New Auto-Interp

Configuration

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

IFrame

Link

Not in Any Lists

Negative Logits

if

-2.02

madeus

-1.69

</b>

-1.56



-1.53

for

-1.51



-1.45

when

-1.43

-1.42

 Przy

-1.38

does

-1.36

POSITIVE LOGITS

各样的

2.75

 arbete

1.88

 pozor

1.80

américa

1.74

 podrob

1.72

reuil

1.70

liances

1.69

 läsa

1.65

 مختلف

1.64

 bruke

1.60

Activations Density 0.012%