INDEX

Explanations

a or the

New Auto-Interp

Configuration

Prompts (Dashboard)

16,384 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

tip

-0.31

-tip

-0.28

Tip

-0.28

 initiative

-0.26

icals

-0.25

 speech

-0.25

inceton

-0.25

vard

-0.25

åĿĢ

-0.24

aseline

-0.24

POSITIVE LOGITS

æĺ¯æľī

0.26

modern

0.26

çİ°ä»£

0.25

èĮ¨

0.25

ãĥªãĤ¢

0.24

ennon

0.24

 prÃ©c

0.24

JOR

0.24

DST

0.23

 flare

0.23

Activations Density 0.105%