INDEX

Explanations

additional context or elements

Simulates or compositional structure

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

 głównie

0.54

 identifications

0.53

 desenvolv

0.46

 cach

0.45

 sont

0.44

 induction

0.44

 identific

0.44

 allong

0.43

 substitutions

0.43

 identificación

0.43

POSITIVE LOGITS

ければ

0.53

hém

0.50

Rank

0.47

се

0.47

Match

0.47

Mit

0.46

Myth

0.46

änglich

0.46

↵↵

0.45

ótes

0.45

Activations Density 0.000%