INDEX

Explanations

No Explanations Found

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

 Extr

-0.11

 PROVIDED

-0.10

ken

-0.10

 poised

-0.10

ouce

-0.10

Hoy

-0.10

/testing

-0.09

ril

-0.09

extr

-0.09

pem

-0.09

POSITIVE LOGITS

vs

0.08

pressed

0.08

vented

0.08

adress

0.08

usters

0.08

 Pref

0.07

 subscript

0.07

ust

0.07

 Rowe

0.07

 myth

0.07

Activations Density 0.000%

No Known Activations

This feature has no known activations.