INDEX

Explanations

comparisons

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

[code

-0.09

attended

-0.09

latent

-0.09

ä¼ļå¼ķèµ·

-0.09

ä¸ºä»Ģä¹Īä¼ļ

-0.09

olland

-0.08

 arrested

-0.08

 ÑĢÐµÐ¹

-0.08

ÑĦÐ¸

-0.08

çıł

-0.08

POSITIVE LOGITS

çĽ¸æ¯Ķä¹ĭä¸ĭ

0.16

 compared

0.15

çĽ¸æ¯Ķ

0.13

 Compared

0.11

çĽ¸å¯¹äºİ

0.11

æ¯Ķèµ·

0.10

å®ıå¤§

0.10

çĻŀ

0.09

çĽ¸æ¯Ķäºİ

0.09

çľ¼åīįçļĦ

0.09

Activations Density 0.044%