INDEX

Explanations

then followed by a consequence

New Auto-Interp

Configuration

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 then

-0.24

then

-0.21

Then

-0.21

	then

-0.18

 Then

-0.18

_then

-0.16

THEN

-0.15

 THEN

-0.15

now

-0.12

 thereafter

-0.12

POSITIVE LOGITS

umber

0.15

ewe

0.11

-current

0.11

ouve

0.11

oner

0.11

moz

0.10

ìĦľ

0.10

 aber

0.10

etwork

0.09

onus

0.09

Activations Density 0.099%