INDEX

Explanations

part of

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

utsch

-0.11

nees

-0.11

 situation

-0.10

-0.09

 Hicks

-0.09

 Walters

-0.09

oble

-0.09

iname

-0.09

RowAt

-0.09

iaux

-0.09

POSITIVE LOGITS

 fabric

0.25

 puzzle

0.22

 equation

0.20

 identity

0.18

fabric

0.16

 makeup

0.16

 Fabric

0.16

 Puzzle

0.16

 overall

0.15

 identities

0.15

Activations Density 0.074%