INDEX

Explanations

need space, time, or a moment

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

 Relax

-0.11

 Webcam

-0.10

 applicable

-0.09

 abandon

-0.09

oco

-0.09

 Fang

-0.09

 relaxation

-0.09

YTE

-0.09

 interpol

-0.09

 Compass

-0.09

POSITIVE LOGITS

 space

0.23

 distance

0.21

space

0.18

 Space

0.18

 Distance

0.17

 SPACE

0.17

è·ĿéĽ¢

0.17

distance

0.17

Distance

0.16

 time

0.16

Activations Density 0.043%