INDEX

Explanations

adjustment or adaptation

New Auto-Interp

Configuration

Prompts (Dashboard)

16,384 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

IFrame

Link

Negative Logits

-0.09

aire

-0.09

 Dani

-0.08

 Result

-0.08

stadt

-0.08

 proudly

-0.08

 transparent

-0.08

occasion

-0.08

 Found

-0.08

kap

-0.07

POSITIVE LOGITS

 adjustment

0.24

 adaptation

0.24

 Adjustment

0.21

 adapting

0.21

ac

0.21

 adjusts

0.21

 Adapt

0.20

 adjust

0.20

 orient

0.20

adjust

0.20

Activations Density 0.125%

No Known Activations