INDEX

Explanations

bank names and code structures

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

if

-1.27

 where

-0.94

 this

-0.91

all

-0.90

at

-0.88

 when

-0.88

دارة

-0.85

 both

-0.84

cipta

-0.84

回到了

-0.83

POSITIVE LOGITS

cách

1.03

ΡΑ

1.00

Примечания

0.98

редиенты

0.93

ią

0.92

 incomplète

0.91

着脸

0.91

 требования

0.91

veria

0.89

gaven

0.89

Activations Density 0.001%