INDEX

Explanations

negation followed by verb phrase

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

ä¸įæĪĲ

-0.30

ä¸įå®¹æĺĵ

-0.27

 zwar

-0.26

ä¸įéĶĪ

-0.26

xmm

-0.25

ä¸įä½Ĩ

-0.25

izza

-0.23

è§£åĨ³éĹ®é¢ĺ

-0.23

çº¹

-0.23

ä¸įåĪĨ

-0.23

POSITIVE LOGITS

é¢ĳç¹ģ

0.24

ä¼ĳ

0.24

å¦ĤæŃ¤

0.23

çĥŃçĪ±

0.23

KEEP

0.23

 redund

0.22

éĹ³

0.22

(Cs

0.22

 constantly

0.22

è¿Ļä¹Īå¤ļ

0.22

Activations Density 0.015%