INDEX

Explanations

original article

New Auto-Interp

Configuration

Prompts (Dashboard)

16,384 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

å¹³æĹ¶

-0.29

æĹ¥å¸¸

-0.26

 particular

-0.24

äºĭåħĪ

-0.24

 immediate

-0.23

åħ·ä½ĵæĥħåĨµ

-0.23

 familiar

-0.23

uat

-0.22

 generally

-0.22

hel

-0.22

POSITIVE LOGITS

/original

0.46

(original

0.43

.original

0.39

 original

0.35

Original

0.34

 Original

0.34

 originally

0.34

original

0.34

ãģ«æĪ»

0.32

 ORIGINAL

0.32

Activations Density 0.089%