INDEX

Explanations

specificity and detail

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

icial

-0.07

WithError

-0.07

Ð¸ÑģÑĤÐ¾

-0.07

éĤ£ä¹Īå¥½

-0.07

 basic

-0.07

ubungan

-0.07

apun

-0.07

åıįæŃ£

-0.07

 apar

-0.07

ä¸ºæĤ¨ä»ĭç»į

-0.07

POSITIVE LOGITS

åħ·ä½ĵçļĦ

0.13

 konkret

0.12

 detail

0.11

 specificity

0.11

è¯¦ç»ĨçļĦ

0.11

 specifics

0.11

ç»ĨèĬĤ

0.11

 details

0.11

 detailing

0.11

åħ·ä½ĵæİªæĸ½

0.11

Activations Density 0.036%