INDEX

Explanations

flexible and permissive

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

Negative Logits

 narrowing

-0.09

alnÄ±z

-0.09

DOMContentLoaded

-0.09

 Pace

-0.09

 WARRANTIES

-0.09

Loy

-0.09

bidden

-0.09

 unauthorized

-0.08

 exclus

-0.08

ä¸įåı¯

-0.08

POSITIVE LOGITS

 relaxed

0.46

len

0.44

 liberal

0.43

 relax

0.42

 relaxation

0.38

lax

0.36

 Relax

0.35

len

0.34

 relaxing

0.31

Len

0.29

Activations Density 0.130%

No Known Activations