Model: GELU Model: 4 Layers, 2048 Neurons per Layer

Dataset: 80% C4 (Web Text) and 20% Python Code

Neuron 133 in Layer 1

Load this data into an Interactive Neuroscope

See Documentation here

Transformer Lens Loading: HookedTransformer.from_pretrained('gelu-4l')

Text #0

Max Range: 3.4004. Min Range: -3.4004

Max Act: 3.4004. Min Act: -0.1700

Data Index: 825712 (C4 (Web Text))

Max Activating Token Index: 51

Text #1

Max Range: 3.4004. Min Range: -3.4004

Max Act: 3.3690. Min Act: -0.1700

Data Index: 711663 (C4 (Web Text))

Max Activating Token Index: 732

Text #2

Max Range: 3.4004. Min Range: -3.4004

Max Act: 3.3690. Min Act: -0.1700

Data Index: 611813 (C4 (Web Text))

Max Activating Token Index: 996

Text #3

Max Range: 3.4004. Min Range: -3.4004

Max Act: 3.3377. Min Act: -0.1700

Data Index: 1087435 (C4 (Web Text))

Max Activating Token Index: 389

Text #4

Max Range: 3.4004. Min Range: -3.4004

Max Act: 3.3062. Min Act: -0.1700

Data Index: 825364 (C4 (Web Text))

Max Activating Token Index: 228

Text #5

Max Range: 3.4004. Min Range: -3.4004

Max Act: 3.2748. Min Act: -0.1700

Data Index: 270802 (C4 (Web Text))

Max Activating Token Index: 893

Text #6

Max Range: 3.4004. Min Range: -3.4004

Max Act: 3.3062. Min Act: -0.1700

Data Index: 646907 (C4 (Web Text))

Max Activating Token Index: 360

Text #7

Max Range: 3.4004. Min Range: -3.4004

Max Act: 3.2748. Min Act: -0.1700

Data Index: 787183 (C4 (Web Text))

Max Activating Token Index: 53

Text #8

Max Range: 3.4004. Min Range: -3.4004

Max Act: 3.3062. Min Act: -0.1700

Data Index: 410183 (C4 (Web Text))

Max Activating Token Index: 510

Text #9

Max Range: 3.4004. Min Range: -3.4004

Max Act: 3.3062. Min Act: -0.1700

Data Index: 1190812 (C4 (Web Text))

Max Activating Token Index: 12

Text #10

Max Range: 3.4004. Min Range: -3.4004

Max Act: 3.2748. Min Act: -0.1700

Data Index: 1154205 (C4 (Web Text))

Max Activating Token Index: 986

Text #11

Max Range: 3.4004. Min Range: -3.4004

Max Act: 3.2748. Min Act: -0.1700

Data Index: 640029 (C4 (Web Text))

Max Activating Token Index: 78

Text #12

Max Range: 3.4004. Min Range: -3.4004

Max Act: 3.3062. Min Act: -0.1700

Data Index: 615156 (C4 (Web Text))

Max Activating Token Index: 861

Text #13

Max Range: 3.4004. Min Range: -3.4004

Max Act: 3.2748. Min Act: -0.1700

Data Index: 690946 (C4 (Web Text))

Max Activating Token Index: 801

Text #14

Max Range: 3.4004. Min Range: -3.4004

Max Act: 3.2434. Min Act: -0.1700

Data Index: 1588004 (Python Code)

Max Activating Token Index: 650

Text #15

Max Range: 3.4004. Min Range: -3.4004

Max Act: 3.2120. Min Act: -0.1700

Data Index: 62693 (C4 (Web Text))

Max Activating Token Index: 199

Text #16

Max Range: 3.4004. Min Range: -3.4004

Max Act: 3.2120. Min Act: -0.1700

Data Index: 776889 (C4 (Web Text))

Max Activating Token Index: 771

Text #17

Max Range: 3.4004. Min Range: -3.4004

Max Act: 3.2120. Min Act: -0.1700

Data Index: 205968 (C4 (Web Text))

Max Activating Token Index: 778

Text #18

Max Range: 3.4004. Min Range: -3.4004

Max Act: 3.2434. Min Act: -0.1700

Data Index: 15785 (C4 (Web Text))

Max Activating Token Index: 653

Text #19

Max Range: 3.4004. Min Range: -3.4004

Max Act: 3.2434. Min Act: -0.1700

Data Index: 1057134 (C4 (Web Text))

Max Activating Token Index: 784