INDEX

Explanations

first week or month

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

 prep

-0.11

 lately

-0.10

 suddenly

-0.10

deg

-0.09

 rejuven

-0.09

ç½

-0.09

 russ

-0.09

 Bryant

-0.09

 recent

-0.08

 afterwards

-0.08

POSITIVE LOGITS

ç¬¬

0.22

 early

0.22

 ç¬¬

0.19

 Ð¿ÐµÑĢÐ²

0.18

 halfway

0.18

early

0.16

 third

0.16

Early

0.15

 first

0.15

 Early

0.15

Activations Density 0.323%