INDEX

Explanations

heavy followed by descriptive nouns

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

 ÑģÐ¾Ð²ÐµÑĢÑĪ

-0.09

marine

-0.09

uve

-0.09

otron

-0.09

afety

-0.09

andre

-0.08

é¹¤

-0.08

æľīå¾Īå¤§

-0.08

æİ¥

-0.08

imated

-0.08

POSITIVE LOGITS

Heavy

0.17

heavy

0.16

 Heavy

0.16

æ²īéĩį

0.16

 heavy

0.15

 heavier

0.13

 náº·ng

0.13

 heav

0.13

-heavy

0.12

åİļéĩį

0.12

Activations Density 0.046%