INDEX

Explanations

Western diet

indexing operations using square brackets in programming code.

New Auto-Interp

Configuration

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

IFrame

Link

Not in Any Lists

Negative Logits

 from

-1.45

↵

-1.44

 only

-1.39

-1.22

从

-1.13

刚

-1.13

for

-1.12

 тільки

-1.11

all

-1.02

 تضيف

-1.02

POSITIVE LOGITS

覿

1.30

__':

1.30



1.28

ível

1.23

iempo

1.21

CIUDAD

1.19

HPP

1.16

 anſ

1.16

étoit

1.16

falfa

1.16

Activations Density 0.004%