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

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

Neuron 392 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: 3.4420. Min Range: -3.4420

Max Act: 3.4420. Min Act: -0.1700

Data Index: 1158618 (C4 (Web Text))

Max Activating Token Index: 641

Text #1

Max Range: 3.4420. Min Range: -3.4420

Max Act: 3.4107. Min Act: -0.1700

Data Index: 1425479 (Python Code)

Max Activating Token Index: 797

Text #2

Max Range: 3.4420. Min Range: -3.4420

Max Act: 3.3793. Min Act: -0.1700

Data Index: 1544934 (Python Code)

Max Activating Token Index: 912

Text #3

Max Range: 3.4420. Min Range: -3.4420

Max Act: 3.3165. Min Act: -0.1700

Data Index: 455857 (C4 (Web Text))

Max Activating Token Index: 776

Text #4

Max Range: 3.4420. Min Range: -3.4420

Max Act: 3.3008. Min Act: -0.1700

Data Index: 1383684 (Python Code)

Max Activating Token Index: 770

Text #5

Max Range: 3.4420. Min Range: -3.4420

Max Act: 3.3008. Min Act: -0.1700

Data Index: 1607375 (Python Code)

Max Activating Token Index: 968

Text #6

Max Range: 3.4420. Min Range: -3.4420

Max Act: 3.3165. Min Act: -0.1700

Data Index: 1651531 (Python Code)

Max Activating Token Index: 1020

Text #7

Max Range: 3.4420. Min Range: -3.4420

Max Act: 3.2537. Min Act: -0.1700

Data Index: 478815 (C4 (Web Text))

Max Activating Token Index: 882

Text #8

Max Range: 3.4420. Min Range: -3.4420

Max Act: 3.3322. Min Act: -0.1700

Data Index: 1454323 (Python Code)

Max Activating Token Index: 760

Text #9

Max Range: 3.4420. Min Range: -3.4420

Max Act: 3.2694. Min Act: -0.1700

Data Index: 1539506 (Python Code)

Max Activating Token Index: 338

Text #10

Max Range: 3.4420. Min Range: -3.4420

Max Act: 3.2851. Min Act: -0.1700

Data Index: 1413967 (Python Code)

Max Activating Token Index: 883

Text #11

Max Range: 3.4420. Min Range: -3.4420

Max Act: 3.2851. Min Act: -0.1700

Data Index: 83363 (C4 (Web Text))

Max Activating Token Index: 651

Text #12

Max Range: 3.4420. Min Range: -3.4420

Max Act: 3.2694. Min Act: -0.1700

Data Index: 844695 (C4 (Web Text))

Max Activating Token Index: 828

Text #13

Max Range: 3.4420. Min Range: -3.4420

Max Act: 3.2694. Min Act: -0.1700

Data Index: 1381155 (Python Code)

Max Activating Token Index: 881

Text #14

Max Range: 3.4420. Min Range: -3.4420

Max Act: 3.3008. Min Act: -0.1700

Data Index: 1481267 (Python Code)

Max Activating Token Index: 933

Text #15

Max Range: 3.4420. Min Range: -3.4420

Max Act: 3.2380. Min Act: -0.1700

Data Index: 1377579 (Python Code)

Max Activating Token Index: 844

Text #16

Max Range: 3.4420. Min Range: -3.4420

Max Act: 3.2537. Min Act: -0.1700

Data Index: 745279 (C4 (Web Text))

Max Activating Token Index: 768

Text #17

Max Range: 3.4420. Min Range: -3.4420

Max Act: 3.2537. Min Act: -0.1700

Data Index: 820603 (C4 (Web Text))

Max Activating Token Index: 278

Text #18

Max Range: 3.4420. Min Range: -3.4420

Max Act: 3.2851. Min Act: -0.1700

Data Index: 1634978 (Python Code)

Max Activating Token Index: 797

Text #19

Max Range: 3.4420. Min Range: -3.4420

Max Act: 3.3008. Min Act: -0.1700

Data Index: 1374686 (Python Code)

Max Activating Token Index: 807