INDEX

Explanations

competition and competing

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

of

-1.45

 densely

-1.08

 interacting

-1.04

 tackling

-1.03

ファッション

-1.00

all

-0.99

 از

-0.97

 abrupt

-0.97

 experimenting

-0.96

 consolidating

-0.95

POSITIVE LOGITS

对手

1.34

鲩

1.14

 Comte

1.13

鯪

1.10

 לקבל

1.09

 côtes

1.07

aggia

1.06

 frontières

1.05

饪

1.03

 Cartel

1.02

Activations Density 0.019%