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

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

Neuron 8 in Layer 1

Load this data into an Interactive Neuroscope

See Documentation here

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

Text #0

Max Range: 3.4869. Min Range: -3.4869

Max Act: 3.4869. Min Act: -0.1700

Data Index: 410701 (C4 (Web Text))

Max Activating Token Index: 7

Text #1

Max Range: 3.4869. Min Range: -3.4869

Max Act: 3.4398. Min Act: -0.1700

Data Index: 915836 (C4 (Web Text))

Max Activating Token Index: 713

Text #2

Max Range: 3.4869. Min Range: -3.4869

Max Act: 3.4085. Min Act: -0.1700

Data Index: 1248840 (C4 (Web Text))

Max Activating Token Index: 55

Text #3

Max Range: 3.4869. Min Range: -3.4869

Max Act: 3.4085. Min Act: -0.1700

Data Index: 582433 (C4 (Web Text))

Max Activating Token Index: 11

Text #4

Max Range: 3.4869. Min Range: -3.4869

Max Act: 3.4085. Min Act: -0.1700

Data Index: 1127466 (C4 (Web Text))

Max Activating Token Index: 532

Text #5

Max Range: 3.4869. Min Range: -3.4869

Max Act: 3.3928. Min Act: -0.1700

Data Index: 1236443 (C4 (Web Text))

Max Activating Token Index: 362

Text #6

Max Range: 3.4869. Min Range: -3.4869

Max Act: 3.3771. Min Act: -0.1700

Data Index: 240893 (C4 (Web Text))

Max Activating Token Index: 717

Text #7

Max Range: 3.4869. Min Range: -3.4869

Max Act: 3.3614. Min Act: -0.1700

Data Index: 74615 (C4 (Web Text))

Max Activating Token Index: 124

Text #8

Max Range: 3.4869. Min Range: -3.4869

Max Act: 3.3457. Min Act: -0.1700

Data Index: 1006598 (C4 (Web Text))

Max Activating Token Index: 861

Text #9

Max Range: 3.4869. Min Range: -3.4869

Max Act: 3.3457. Min Act: -0.1700

Data Index: 1180388 (C4 (Web Text))

Max Activating Token Index: 973

Text #10

Max Range: 3.4869. Min Range: -3.4869

Max Act: 3.2986. Min Act: -0.1700

Data Index: 1233511 (C4 (Web Text))

Max Activating Token Index: 732

Text #11

Max Range: 3.4869. Min Range: -3.4869

Max Act: 3.2986. Min Act: -0.1700

Data Index: 1031081 (C4 (Web Text))

Max Activating Token Index: 543

Text #12

Max Range: 3.4869. Min Range: -3.4869

Max Act: 3.2986. Min Act: -0.1700

Data Index: 45230 (C4 (Web Text))

Max Activating Token Index: 1008

Text #13

Max Range: 3.4869. Min Range: -3.4869

Max Act: 3.2672. Min Act: -0.1700

Data Index: 813255 (C4 (Web Text))

Max Activating Token Index: 345

Text #14

Max Range: 3.4869. Min Range: -3.4869

Max Act: 3.2829. Min Act: -0.1700

Data Index: 1146912 (C4 (Web Text))

Max Activating Token Index: 336

Text #15

Max Range: 3.4869. Min Range: -3.4869

Max Act: 3.2672. Min Act: -0.1700

Data Index: 50690 (C4 (Web Text))

Max Activating Token Index: 313

Text #16

Max Range: 3.4869. Min Range: -3.4869

Max Act: 3.2672. Min Act: -0.1700

Data Index: 791496 (C4 (Web Text))

Max Activating Token Index: 707

Text #17

Max Range: 3.4869. Min Range: -3.4869

Max Act: 3.2358. Min Act: -0.1700

Data Index: 1149581 (C4 (Web Text))

Max Activating Token Index: 511

Text #18

Max Range: 3.4869. Min Range: -3.4869

Max Act: 3.2515. Min Act: -0.1700

Data Index: 1221242 (C4 (Web Text))

Max Activating Token Index: 33

Text #19

Max Range: 3.4869. Min Range: -3.4869

Max Act: 3.2358. Min Act: -0.1700

Data Index: 1333367 (C4 (Web Text))

Max Activating Token Index: 422