INDEX

Explanations

first name, last name

This neuron fires on occurrences of the “first name” field label (i.e. tokens like “First” + “name” in forms, code, or data schemas).

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

on

-1.86

</u>

-1.73

</b>

-1.52

更好

-1.46

</em>

-1.46

ches

-1.45

sed

-1.42

…

-1.41

-1.36

for

-1.34

POSITIVE LOGITS

Inconvénients

1.63

賈

1.59

邊

1.59

 energías

1.55

 partículas

1.55

 expérimentés

1.52

Both

1.49

 正面

1.43

rostis

1.43

 adhesivo

1.42

Activations Density 0.030%