INDEX

Explanations

for the

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

</em>

-1.33

 случае

-1.12

 hlavně

-1.05

現在は

-1.04

 antigüedad

-1.03

 они

-1.02

 with

-1.00

 вашего

-1.00

 клиент

-0.97

POSITIVE LOGITS

 what

1.34

 something

1.21

 דבר

1.19

 véhicule

1.16

égard

1.13

 emite

1.12

 anything

1.11

 akan

1.07

 ukon

1.07

 veicolo

1.06

Activations Density 0.042%