INDEX

Explanations

increase or worsening

New Auto-Interp

Configuration

Prompts (Dashboard)

16,384 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 Razor

-0.09

UNU

-0.09

 hopefully

-0.09

 outright

-0.09

 maxim

-0.09

 Trem

-0.09

Hob

-0.08

 Const

-0.08

imar

-0.08

Rud

-0.08

POSITIVE LOGITS

 more

0.14

 MORE

0.14

æĽ´

0.12

 desire

0.12

-more

0.12

 thirst

0.11

 faster

0.11

 harder

0.11

 determination

0.11

efa

0.11

Activations Density 0.025%