INDEX

Explanations

seeking clarification or pointing out errors

New Auto-Interp

Configuration

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

푉

-1.03

 hilfreich

-1.02

bry

-0.98

 nötig

-0.93

 poda

-0.92

 rêves

-0.92

ΗΣ

-0.91

DEN

-0.91

 vincere

-0.89

Eres

-0.89

POSITIVE LOGITS

but

1.25

or

1.23

 this

1.14

 there

1.02

all

1.01

或是

1.00

so

0.91

illeurs

0.89

 allemaal

0.88

manfaat

0.87

Activations Density 0.039%