INDEX

Explanations

building the

The neuron is keyed to the initial tokens of new sentences or clauses—discourse-opening words or punctuation that mark the start of a sentence or parenthetical.

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

」,

-1.32

 недавно

-1.32

Ṇ

-1.28

🔻

-1.26

 Ofic

-1.26

 Перейти

-1.23

 garg

-1.23

мся

-1.23

्ञ

-1.22

 expre

-1.21

POSITIVE LOGITS

It

1.53

it

1.46

all

1.44

If

1.32

Britain

1.30

You

1.26

 only

1.24

of

1.24

Which

1.20

 nekem

1.18

Activations Density 0.007%