INDEX

Explanations

him and context

actions followed by completions

New Auto-Interp

Configuration

Prompts (Dashboard)

392,802 prompts, 256 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

IFrame

Link

Negative Logits

ARIES

0.43

곤

0.42

덥

0.41

ре

0.40

を見る

0.40

 poils

0.40

 nuova

0.39

 oreilles

0.39

 occup

0.39

요

0.39

POSITIVE LOGITS

 sheaf

0.50

 DENUMIRE

0.48

有一定的

0.47

ährend

0.46

ţ

0.46

ombasa

0.45

rás

0.45

 กัน

0.45

 butane

0.44

éfonos

0.44

Activations Density 0.001%