INDEX

Explanations

vegetables and fruits

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

 é«

-0.09

rais

-0.09

áº»

-0.09

meal

-0.09

 Grape

-0.09

 Seed

-0.08

.createObject

-0.08

oso

-0.08

POSIT

-0.08

Meal

-0.08

POSITIVE LOGITS

 vegetables

0.46

 veggies

0.36

èĶ

0.31

 vegetable

0.29

 fruits

0.28

veget

0.27

eggies

0.25

 Veget

0.25

veg

0.24

 greens

0.23

Activations Density 0.075%