Model: GELU Model: 3 Layers, 2048 Neurons per Layer

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

Neuron 66 in Layer 2

Load this data into an Interactive Neuroscope

See Documentation here

Transformer Lens Loading: HookedTransformer.from_pretrained('gelu-3l')

Text #0

Max Range: 3.9024. Min Range: -3.9024

Max Act: 3.9024. Min Act: -0.1700

Data Index: 698173 (C4 (Web Text))

Max Activating Token Index: 932

Text #1

Max Range: 3.9024. Min Range: -3.9024

Max Act: 3.8085. Min Act: -0.1700

Data Index: 1007553 (C4 (Web Text))

Max Activating Token Index: 637

Text #2

Max Range: 3.9024. Min Range: -3.9024

Max Act: 3.7460. Min Act: -0.1700

Data Index: 128239 (C4 (Web Text))

Max Activating Token Index: 483

Text #3

Max Range: 3.9024. Min Range: -3.9024

Max Act: 3.7773. Min Act: -0.1700

Data Index: 1191503 (C4 (Web Text))

Max Activating Token Index: 574

Text #4

Max Range: 3.9024. Min Range: -3.9024

Max Act: 3.6834. Min Act: -0.1700

Data Index: 700540 (C4 (Web Text))

Max Activating Token Index: 875

Text #5

Max Range: 3.9024. Min Range: -3.9024

Max Act: 3.6834. Min Act: -0.1700

Data Index: 1166246 (C4 (Web Text))

Max Activating Token Index: 274

Text #6

Max Range: 3.9024. Min Range: -3.9024

Max Act: 3.7147. Min Act: -0.1700

Data Index: 675530 (C4 (Web Text))

Max Activating Token Index: 900

Text #7

Max Range: 3.9024. Min Range: -3.9024

Max Act: 3.6834. Min Act: -0.1700

Data Index: 1524775 (Python Code)

Max Activating Token Index: 299

Text #8

Max Range: 3.9024. Min Range: -3.9024

Max Act: 3.6834. Min Act: -0.1700

Data Index: 1185950 (C4 (Web Text))

Max Activating Token Index: 856

Text #9

Max Range: 3.9024. Min Range: -3.9024

Max Act: 3.6834. Min Act: -0.1700

Data Index: 551677 (C4 (Web Text))

Max Activating Token Index: 903

Text #10

Max Range: 3.9024. Min Range: -3.9024

Max Act: 3.6208. Min Act: -0.1700

Data Index: 726891 (C4 (Web Text))

Max Activating Token Index: 709

Text #11

Max Range: 3.9024. Min Range: -3.9024

Max Act: 3.5895. Min Act: -0.1700

Data Index: 848288 (C4 (Web Text))

Max Activating Token Index: 510

Text #12

Max Range: 3.9024. Min Range: -3.9024

Max Act: 3.6521. Min Act: -0.1700

Data Index: 319120 (C4 (Web Text))

Max Activating Token Index: 979

Text #13

Max Range: 3.9024. Min Range: -3.9024

Max Act: 3.5895. Min Act: -0.1700

Data Index: 282510 (C4 (Web Text))

Max Activating Token Index: 586

Text #14

Max Range: 3.9024. Min Range: -3.9024

Max Act: 3.6208. Min Act: -0.1700

Data Index: 150462 (C4 (Web Text))

Max Activating Token Index: 813

Text #15

Max Range: 3.9024. Min Range: -3.9024

Max Act: 3.5895. Min Act: -0.1700

Data Index: 381536 (C4 (Web Text))

Max Activating Token Index: 782

Text #16

Max Range: 3.9024. Min Range: -3.9024

Max Act: 3.6208. Min Act: -0.1700

Data Index: 935596 (C4 (Web Text))

Max Activating Token Index: 534

Text #17

Max Range: 3.9024. Min Range: -3.9024

Max Act: 3.5581. Min Act: -0.1700

Data Index: 973048 (C4 (Web Text))

Max Activating Token Index: 707

Text #18

Max Range: 3.9024. Min Range: -3.9024

Max Act: 3.5581. Min Act: -0.1700

Data Index: 797785 (C4 (Web Text))

Max Activating Token Index: 720

Text #19

Max Range: 3.9024. Min Range: -3.9024

Max Act: 3.5581. Min Act: -0.1700

Data Index: 1003901 (C4 (Web Text))

Max Activating Token Index: 598