INDEX

Explanations

minimum values and limits

The neuron activates on numeric literals (e.g. numbers, especially those labeling “min” or other numeric values).

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

 sogenannten

-1.41

 forego

-1.32

hline

-1.25

𝓈

-1.17

max

-1.16

我们

-1.13

 soar

-1.13

Skocz

-1.12

 целей

-1.11

-1.10

POSITIVE LOGITS

 erkannt

1.29

orbit

1.18

 messi

1.16

ȅ

1.16

اته

1.16

 intenta

1.15

ximity

1.14

 atleast

1.13

cams

1.13

chemin

1.11

Activations Density 0.035%