INDEX

Explanations

identifying descriptive statements

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

as

-1.07

これは

-1.04

この

-1.01

”

-0.98

 this

-0.96

</strong>

-0.94

 with

-0.93

Ｓ

-0.91

𝐨

-0.90

ピ

-0.90

POSITIVE LOGITS

exemple

1.23

ということを

1.17

 takich

1.16

 mówią

1.14

 such

1.13

jshint

1.09

 parlent

1.04

筈

1.02

 lantaran

1.02

 yalnızca

1.02

Activations Density 0.040%