INDEX

Explanations

possibilities and hypotheses

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

çŀ»

-0.07

ä¹Łå°±

-0.06

ä¸įä»ħ

-0.06

åģ¥

-0.06

å³ł

-0.06

èĻĶ

-0.06

iscos

-0.06

áº¥m

-0.06

isht

-0.06

éķ¿

-0.06

POSITIVE LOGITS

 interpretations

0.11

 possibilities

0.10

 possible

0.10

åı¦ä¸Ģç§į

0.10

ä¹Łæľīåı¯èĥ½

0.10

æľīåı¯èĥ½æĺ¯

0.09

 scenarios

0.09

 possibility

0.09

 interpretation

0.09

 scenario

0.09

Activations Density 0.025%