INDEX

Explanations

screaming, shouting, yelling

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

 smirk

-0.16

 subtly

-0.14

 passive

-0.14

 quietly

-0.13

 sniff

-0.13

 sinister

-0.12

 smiled

-0.12

 Passive

-0.12

 patiently

-0.12

 silently

-0.12

POSITIVE LOGITS

 shouting

0.44

 yelling

0.41

 shout

0.38

 shouts

0.37

 screaming

0.36

 shouted

0.35

 yell

0.34

 scream

0.34

 yelled

0.33

 screams

0.30

Activations Density 0.140%