INDEX

Explanations

rearview mirror, cameras, or vision

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

scr

-0.09

 dirt

-0.08

wr

-0.08

Wor

-0.08

ud

-0.08

 colle

-0.08

sty

-0.08

 Curl

-0.08

vel

-0.08

gua

-0.08

POSITIVE LOGITS

 lenses

0.15

 lens

0.13

 mirror

0.12

 mirrors

0.12

mirror

0.11

 view

0.11

yon

0.10

 vision

0.10

ing

0.10

roscope

0.10

Activations Density 0.001%