INDEX

Explanations

vasodilation and muscle relaxation

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

ãĢĢï½Į

-0.09

.Unity

-0.09

 turnout

-0.09

 Gardner

-0.09

AAF

-0.09

OLON

-0.09

PLIC

-0.09

 Tham

-0.08

eor

-0.08

 tieten

-0.08

POSITIVE LOGITS

dil

0.30

 dilation

0.29

vas

0.29

vas

0.25

 relax

0.24

Vas

0.24

Dil

0.23

 relaxation

0.23

 relaxing

0.21

 Relax

0.19

Activations Density 0.063%