INDEX

Explanations

app, PH, RemoteException, Baseball, energy

The neuron strongly activates on content‐heavy, domain-specific or technical terms (e.g. specialized jargon and proper nouns) rather than on common function words.

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

tése

-1.38

irée

-1.34

ściu

-1.30

関する

-1.30

catcher

-1.26

 его

-1.25

ógicos

-1.25

腕

-1.24

ashed

-1.22

IFESTYLE

-1.22

POSITIVE LOGITS

 which

1.69

to

1.65

 only

1.48

 would

1.41

 from

1.39

ichtete

1.34

 some

1.33

 place

1.32

for

1.27

Activations Density 0.280%