INDEX

Explanations

vehicle crashes and accidents

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

 Injury

-0.13

edor

-0.12

idot

-0.12

inj

-0.11

akan

-0.11

ivel

-0.11

vard

-0.11

 boots

-0.11

oise

-0.11

 reperc

-0.10

POSITIVE LOGITS

 obliv

0.09

unb

0.09

 someone

0.09

sob

0.08

 enter

0.08

 entering

0.08

 contest

0.08

 marketing

0.08

oug

0.08

Activations Density 0.025%