INDEX

Explanations

appearance and attire

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

 sometimes

-1.48

 Sometimes

-1.27

 sempre

-1.23

 often

-1.23

 πάντα

-1.16

 after

-1.14

 Often

-1.12

 always

-1.10

 during

-1.09

Sometimes

-1.06

POSITIVE LOGITS

 wore

1.46

Judging

1.18

 wearing

1.18

 judging

1.16

Wearing

1.09

 noticing

1.04

seemed

1.00

 dressed

0.99

 estavam

0.98

 Judging

0.97

Activations Density 0.018%