INDEX

Explanations

components and matrix

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

hlen

-0.07

ÏĥÏĦÎ¬

-0.07

_subtype

-0.07

enga

-0.07

é©¬åĬĽ

-0.07

 ajan

-0.07

èķĥ

-0.07

é£ŀåīĳ

-0.07

ÏĥÏĦÎ±

-0.06

(pair

-0.06

POSITIVE LOGITS

 originally

0.12

åİŁæľ¬

0.12

åİŁæĿ¥

0.11

æľ¬å°±

0.11

åİŁæľ¬æĺ¯

0.11

æľ¬æĿ¥å°±

0.11

 Originally

0.10

æľ¬æĿ¥

0.10

åİŁæĿ¥æĺ¯

0.10

à¸Ńà¸¢à¸¹à¹Īà¹ģà¸¥à¹īà¸§

0.10

Activations Density 0.330%