INDEX

Explanations

against government snoops

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

 will

-1.66

 having

-1.45

 with

-1.39

for

-1.25

ׇ

-1.20

but

-1.19

 there

-1.17

取締

-1.12

 obviously

-1.12

 excellent

-1.10

POSITIVE LOGITS

 felicitación

1.16

 useAuth

1.15

 altavoz

1.11

 استناد

1.09

hri

1.09

уд

1.09

 spodnie

1.08

illige

1.07

 ที่

1.07

exemple

1.05

Activations Density 0.089%