INDEX

Explanations

suddenly followed by a word

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

 Surprise

-0.11

 aston

-0.10

 shocked

-0.10

 surprise

-0.10

 scream

-0.10

 surprised

-0.10

 screams

-0.10

 Shock

-0.09

Ced

-0.09

 astonished

-0.09

POSITIVE LOGITS

 VERY

0.11

 weary

0.11

 wish

0.10

 wishing

0.10

 thirsty

0.10

 desirable

0.10

 palms

0.10

 tired

0.10

 Sanchez

0.10

 hungry

0.09

Activations Density 0.025%