INDEX

Explanations

doing it yourself

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

["$

-0.20

uhe

-0.19

raham

-0.19

arkan

-0.19

{name

-0.19

 *----------------------------------------------------------------

-0.18

tps

-0.18

æĹłè¾ľ

-0.18

 Dems

-0.18

-----------*/↵

-0.18

POSITIVE LOGITS

èĩªè¡Į

0.90

èĩªä¸»

0.82

 independently

0.58

 autonom

0.55

 èĩª

0.54

own

0.54

èĩªå·±

0.54

èĩªçĲĨ

0.54

èĩª

0.53

èĩªåĪĨãģ§

0.53

Activations Density 0.912%