INDEX

Explanations

colleagues, peers, contemporaries

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

oland

-0.10

æŁıæŀĹ

-0.10

ä¸¤å®¶

-0.09

å±ķæľĽ

-0.09

æł¹æľ¬æ²¡æľī

-0.09

asto

-0.09

çļĩå®¶

-0.09

èºº

-0.09

ä¸¤ä¸ªäºº

-0.08

æĴ¤

-0.08

POSITIVE LOGITS

 peers

0.16

 peer

0.16

 fellow

0.15

peer

0.13

Peer

0.13

 Peer

0.13

æľĭ

0.12

(peer

0.11

.peer

0.11

-peer

0.11

Activations Density 0.063%