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

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

Neuron 900 in Layer 3

Load this data into an Interactive Neuroscope

See Documentation here

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

Text #0

Max Range: 5.3665. Min Range: -5.3665

Max Act: 5.3665. Min Act: -0.1700

Data Index: 593861 (C4 (Web Text))

Max Activating Token Index: 286

Text #1

Max Range: 5.3665. Min Range: -5.3665

Max Act: 5.2415. Min Act: -0.1700

Data Index: 147403 (C4 (Web Text))

Max Activating Token Index: 831

Text #2

Max Range: 5.3665. Min Range: -5.3665

Max Act: 5.2415. Min Act: -0.1700

Data Index: 1470569 (Python Code)

Max Activating Token Index: 987

Text #3

Max Range: 5.3665. Min Range: -5.3665

Max Act: 5.2415. Min Act: -0.1700

Data Index: 811200 (C4 (Web Text))

Max Activating Token Index: 149

Text #4

Max Range: 5.3665. Min Range: -5.3665

Max Act: 5.1790. Min Act: -0.1700

Data Index: 502127 (C4 (Web Text))

Max Activating Token Index: 534

Text #5

Max Range: 5.3665. Min Range: -5.3665

Max Act: 5.2728. Min Act: -0.1700

Data Index: 1578084 (Python Code)

Max Activating Token Index: 955

Text #6

Max Range: 5.3665. Min Range: -5.3665

Max Act: 5.2103. Min Act: -0.1700

Data Index: 1196963 (C4 (Web Text))

Max Activating Token Index: 587

Text #7

Max Range: 5.3665. Min Range: -5.3665

Max Act: 5.1790. Min Act: -0.1700

Data Index: 450897 (C4 (Web Text))

Max Activating Token Index: 465

Text #8

Max Range: 5.3665. Min Range: -5.3665

Max Act: 5.1790. Min Act: -0.1700

Data Index: 749819 (C4 (Web Text))

Max Activating Token Index: 342

Text #9

Max Range: 5.3665. Min Range: -5.3665

Max Act: 5.2103. Min Act: -0.1700

Data Index: 1235507 (C4 (Web Text))

Max Activating Token Index: 796

Text #10

Max Range: 5.3665. Min Range: -5.3665

Max Act: 5.1790. Min Act: -0.1700

Data Index: 854230 (C4 (Web Text))

Max Activating Token Index: 814

Text #11

Max Range: 5.3665. Min Range: -5.3665

Max Act: 5.2103. Min Act: -0.1700

Data Index: 21957 (C4 (Web Text))

Max Activating Token Index: 786

Text #12

Max Range: 5.3665. Min Range: -5.3665

Max Act: 5.1790. Min Act: -0.1700

Data Index: 1154008 (C4 (Web Text))

Max Activating Token Index: 631

Text #13

Max Range: 5.3665. Min Range: -5.3665

Max Act: 5.1790. Min Act: -0.1700

Data Index: 1294879 (C4 (Web Text))

Max Activating Token Index: 612

Text #14

Max Range: 5.3665. Min Range: -5.3665

Max Act: 5.1478. Min Act: -0.1700

Data Index: 1102167 (C4 (Web Text))

Max Activating Token Index: 792

Text #15

Max Range: 5.3665. Min Range: -5.3665

Max Act: 5.1478. Min Act: -0.1700

Data Index: 155909 (C4 (Web Text))

Max Activating Token Index: 515

Text #16

Max Range: 5.3665. Min Range: -5.3665

Max Act: 5.1165. Min Act: -0.1700

Data Index: 1220184 (C4 (Web Text))

Max Activating Token Index: 264

Text #17

Max Range: 5.3665. Min Range: -5.3665

Max Act: 5.1165. Min Act: -0.1700

Data Index: 1220343 (C4 (Web Text))

Max Activating Token Index: 472

Text #18

Max Range: 5.3665. Min Range: -5.3665

Max Act: 5.1165. Min Act: -0.1700

Data Index: 29532 (C4 (Web Text))

Max Activating Token Index: 150

Text #19

Max Range: 5.3665. Min Range: -5.3665

Max Act: 5.1478. Min Act: -0.1700

Data Index: 416598 (C4 (Web Text))

Max Activating Token Index: 316