INDEX

Explanations

AD or Ad at the start of words

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

vertisement

-0.12

Advertisement

-0.11

ero

-0.10

vertise

-0.10

Advertis

-0.10

illation

-0.09

 appreh

-0.09

eros

-0.09

commons

-0.09

 Rahul

-0.09

POSITIVE LOGITS

 nause

0.17

lib

0.13

obe

0.13

ored

0.12

orners

0.12

resses

0.12

orno

0.12

ju

0.12

ams

0.12

esso

0.11

Activations Density 0.031%