INDEX

Explanations

negative constructions and phrases indicating exceptions or contradictions

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

seau

-0.07

amburger

-0.06

.remote

-0.06

aux

-0.06

 MetroFramework

-0.06

EGIN

-0.06

tek

-0.06

xb

-0.06

textInput

-0.06

ida

-0.06

POSITIVE LOGITS

emain

0.07

ACLE

0.07

orian

0.07

inton

0.07

sing

0.07

ptest

0.06

 Reform

0.06

ependency

0.06

ixo

0.06

 reform

0.06

Activations Density 0.010%