INDEX

Explanations

calculate rates, scores, and hypotheses

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

ján

-1.05

 arbejde

-1.05

 suele

-1.05

 النبات

-1.04

きましたが

-0.98

ᱸ

-0.98

 WITNESS

-0.97

larınız

-0.96

 liée

-0.96

tais

-0.95

POSITIVE LOGITS

for

1.51

"",

1.17

how

1.17

 updated

1.10

 each

1.10

्यालय

1.07

 لكل

1.06

 października

1.04

 which

1.02

 corresponding

1.02

Activations Density 0.101%