INDEX

Explanations

saying nothing

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

deg

-0.09

 Demp

-0.08

 syndrome

-0.08

 hare

-0.07

cline

-0.07

 brain

-0.07

ews

-0.07

 flat

-0.07

 mart

-0.07

-dis

-0.06

POSITIVE LOGITS

tac

0.31

 implicit

0.31

 silently

0.30

 silent

0.29

spoken

0.28

 spoken

0.28

 communicated

0.26

uns

0.26

 implicitly

0.25

 Implicit

0.24

Activations Density 0.070%