INDEX

Explanations

client or customer followed by colon

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

askan

-0.08

bery

-0.08

empt

-0.08

dens

-0.08

/Foundation

-0.08

abras

-0.08

.UnitTesting

-0.08

_TBL

-0.08

Ã¢k

-0.08

CJK

-0.08

POSITIVE LOGITS

0.15

::

0.12

 Conway

0.10

âĢ²

0.10

 brick

0.09

Tut

0.09

NP

0.09

 Brick

0.08

 Crom

0.08

bul

0.08

Activations Density 0.019%