INDEX

Explanations

code declarations and definitions

The neuron fires on tokens that begin with a capital letter—i.e. uppercase-initial identifiers and type names.

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

 first

-1.74

 himo

-1.41

(^^

-1.41

cended

-1.41

 domestique

-1.38

 …)

-1.38

 dava

-1.36

 Certo

-1.34

__('

-1.34

 deras

-1.32

POSITIVE LOGITS

of

2.20

',

1.67

1.51

 será

1.49

 zelfs

1.47

1.45

",

1.43

 that

1.34

 itulah

1.34

却又

1.34

Activations Density 0.036%