INDEX

Explanations

marathon running and events

New Auto-Interp

Configuration

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 There

-2.42

 unſ

-2.25

出来て

-2.22

 THERE

-2.19

大多

-2.19

に行って

-2.16

 Therefore

-2.11

 पूरी

-2.08

 และ

-2.05

 However

-2.02

POSITIVE LOGITS

3.17

2.86

های

2.31

frac

2.30

2.16

2.11

숱

2.08

臜

2.05

 ravi

2.05

2.03

Activations Density 0.005%