INDEX

Explanations

expressions of emotional conflict or ambiguity

New Auto-Interp

Configuration

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

ĳ

-0.07

heed

-0.06

 pretending

-0.06

eldom

-0.06

erap

-0.06

alth

-0.06

 nonsense

-0.06

 pretend

-0.06

onder

-0.06

tod

-0.06

POSITIVE LOGITS

 sound

0.09

sound

0.08

_SOUND

0.08

Sound

0.08

 sounds

0.08

 SOUND

0.08

zik

0.08

 sounding

0.08

 Sound

0.07

_sound

0.07

Activations Density 0.008%