INDEX

Explanations

usernames or names

The neuron fires on schema or code identifiers—i.e. variable, field, or property names (like “firstname,” “lastname,” “Username,” or the “complete” in “autocomplete”).

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

齶

-2.42

-2.31

 ważne

-2.30

 márm

-2.16

}{

-2.08

‌ای

-2.06

ähig

-2.02

("

-2.02

</u>

-2.00

込んだ

-1.99

POSITIVE LOGITS

you

2.53

not

2.47

ها

2.41

2.34

 disseminating

2.33

鄠

2.27

嘡

2.25

賔

2.23

 поэтому

2.23

as

2.23

Activations Density 0.019%