INDEX

Explanations

named or called

New Auto-Interp

Configuration

Prompts (Dashboard)

16,384 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

/at

-0.11

 tung

-0.10

_readable

-0.10

 disb

-0.10

 Girlfriend

-0.09

cust

-0.09

 originals

-0.09

greg

-0.09

mpl

-0.09

 childcare

-0.09

POSITIVE LOGITS

 named

0.18

Called

0.15

named

0.14

 Named

0.13

 called

0.13

 Called

0.13

Named

0.12

 whom

0.12

called

0.12

 unnamed

0.12

Activations Density 0.120%