INDEX

Explanations

permission and authorization

New Auto-Interp

Top Features by Cosine Similarity

Configuration

Prompts (Dashboard)

10,000 prompts, 128 tokens each

Dataset (Dashboard)

lmsys/lmsys-chat-1m

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

------+------+

-0.09

LOCKS

-0.09

.scalablytyped

-0.09

:Register

-0.08

ban

-0.08

 Advisory

-0.08

 Perf

-0.08

 outs

-0.08

 memberships

-0.08

 Guaranteed

-0.08

POSITIVE LOGITS

 permission

0.64

 Permission

0.50

permission

0.49

Permission

0.45

 permissions

0.45

_permission

0.36

.permission

0.35

 PERMISSION

0.35

æĿĥéĻĲ

0.34

 izin

0.34

Activations Density 0.122%