INDEX

Explanations

The neuron activates primarily on numeric character sequences (i.e. numeric literals) in the text.

SAP GUI, function, map

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

 create

-1.41

at

-1.40

-1.38

all

-1.36

an

-1.29

you

-1.27

 best

-1.26

↵↵

-1.18

for

-1.17

 Best

-1.17

POSITIVE LOGITS

⾮

1.43

 informaciones

1.37

УДК

1.37

价比

1.36

Contexte

1.34



1.34

也非常

1.30

Varios

1.28

 traducciones

1.28

会出现

1.27

Activations Density 0.000%