INDEX

Explanations

options and choices

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

bridge

-0.09

itte

-0.09

ParameterValue

-0.09

éĽ¶

-0.08

 Pierce

-0.08

 avenue

-0.08

ForKey

-0.08

ovit

-0.08

 Selected

-0.08

POSITIVE LOGITS

 choices

0.44

 options

0.42

 choice

0.34

 Options

0.34

options

0.33

choices

0.32

 Choices

0.30

Options

0.29

 opciones

0.28

.options

0.27

Activations Density 0.211%