INDEX

Explanations

right followed by direction

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

以上

-0.80

AGAIN

-0.76

 novamente

-0.75

 AUTHORITIES

-0.73

宇宙

-0.73

 تھا

-0.72

Gibbs

-0.72

 AGAIN

-0.72

isamment

-0.71

bür

-0.71

POSITIVE LOGITS

 right

4.13

right

2.95

Right

2.66

 Right

2.36

RIGHT

2.13

 RIGHT

1.96

 straight

1.66

 ngay

1.60

 derecho

1.56

 derecha

1.55

Activations Density 0.036%