INDEX

Explanations

keep an eye on

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

 visibility

-0.09

 actionTypes

-0.09

 Visibility

-0.08

 Viewing

-0.08

YLE

-0.08

scr

-0.08

Gow

-0.08

onth

-0.08

 Nimbus

-0.08

ãĤĽ

-0.08

POSITIVE LOGITS

 watch

0.43

eye

0.40

eye

0.33

Eye

0.31

watch

0.31

 Watch

0.30

-eye

0.30

Eye

0.29

.watch

0.28

Watch

0.28

Activations Density 0.033%