INDEX

Explanations

Coober Pedy, LNG terminal, Peru

This neuron activates on subword segments that form proper names and named entities (e.g. drug names, organizations, place 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

 nabízí

-1.50

 umožňuje

-1.48

かなぁ

-1.48

 décider

-1.46

 avem

-1.45

 again

-1.41

 opět

-1.41

薷

-1.38

 dovr

-1.34

 réserver

-1.34

POSITIVE LOGITS

 that

2.13

for

1.94

or

1.92

of

1.78

SupportActionBar

1.55

 که

1.52

fordshire

1.51

that

1.48

 یا

1.41

 или

1.35

Activations Density 0.319%