INDEX

Explanations

plus a

Harmful Intent: Analysis

New Auto-Interp

Configuration

Prompts (Dashboard)

392,802 prompts, 256 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

IFrame

Link

Not in Any Lists

Negative Logits

 naturais

0.46

璞

0.46

чні

0.46

瓷

0.45

 earthenware

0.45

ုပ်

0.44

 naturelles

0.44

น

0.43

܂

0.43

دة

0.42

POSITIVE LOGITS

aps

0.56

Analysis

0.45

 Systems

0.43

appa

0.42

oca

0.42

 Analysis

0.42

ile

0.41

 Switches

0.41

wrong

0.40

 Regression

0.40

Activations Density 0.000%