INDEX

Explanations

avoid generating harmful content

New Auto-Interp

Configuration

Prompts (Dashboard)

238,145 prompts, 512 tokens each

Dataset (Dashboard)

lmsys + oasst1

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 ischa

1.20

 здоровья

1.14

쫌

1.09

 związ

1.08

 aprov

1.03

傈

1.03

 ischemic

1.02

 полномо

1.01

й

1.00

 څنګه

0.96

POSITIVE LOGITS

 pitfalls

1.48

諱

1.35

ের

1.25

τητας

1.19

نك

1.16

på

1.14

ين

1.13

onc

1.13

traf

1.12

Activations Density 0.027%