possession and expression

New Auto-Interp

Configuration

Prompts (Dashboard)

16,384 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

IFrame

Link

Negative Logits

osal

-0.10

ego

-0.09

bote

-0.09

ainter

-0.09

 intros

-0.08

 Vienna

-0.08

uru

-0.08

 contempl

-0.08

biz

-0.08

POSITIVE LOGITS

 silence

0.16

 words

0.13

 posture

0.13

 tone

0.12

 visuals

0.11

 language

0.11

 sheer

0.10

 stare

0.10

 Silence

0.10

 silent

0.10

Activations Density 0.035%

No Known Activations