INDEX

Explanations

future possibility or advice

Here's a breakdown

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

饭

0.37

𝐢

0.37

 Gleich

0.36

ミルク

0.35

)?;

0.34

%%%%%%%%%%%%%

0.34

('-');

0.34

珺

0.34

はじめ

0.34

 Weir

0.34

POSITIVE LOGITS

nk

0.48

 arreg

0.46

 scheda

0.44

 cres

0.42

τες

0.42

 रेसि

0.40

বিত

0.39

 에게

0.38

 arguments

0.38

Tk

0.38

Activations Density 0.016%