INDEX

Explanations

verbs introducing claims or arguments

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

läge

-1.00

)}

-0.99

 Before

-0.94

レスレット

-0.94

 Following

-0.92

ského

-0.92

 "-",

-0.91

 what

-0.91

and

-0.90

 ساتھ

-0.90

POSITIVE LOGITS

 they

1.43

 that

1.15

 pensent

1.13

 слишком

1.09

приклад

1.06

 powin

1.05

 deberían

1.04

 przecież

1.02

 cần

1.01

 permitan

0.98

Activations Density 0.026%