INDEX

Explanations

weight and support

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

 underground

-0.13

 Underground

-0.12

 underwater

-0.12

 Escort

-0.11

 escort

-0.11

 Whip

-0.10

 escorts

-0.10

elow

-0.10

 escorted

-0.10

 unearth

-0.10

POSITIVE LOGITS

 weight

0.23

 Weight

0.20

weight

0.20

 weights

0.20

Weight

0.19

 heavier

0.18

-weight

0.18

 weigh

0.17

 weighs

0.16

_weight

0.16

Activations Density 0.088%