INDEX

Explanations

new variations or extensions

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

ODO

-0.10

itore

-0.10

 Changing

-0.10

åĨ

-0.09

åĩĨ

-0.09

_UTF

-0.09

ktop

-0.08

Changing

-0.08

/***/

-0.08

 Bris

-0.08

POSITIVE LOGITS

 variant

0.47

 variation

0.43

 variants

0.40

 variations

0.37

 version

0.36

variant

0.35

 extension

0.33

 Variant

0.31

 Variation

0.31

 extensions

0.30

Activations Density 0.261%