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

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

Neuron 1485 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: 4.1274. Min Range: -4.1274

Max Act: 4.1274. Min Act: -0.1700

Data Index: 788336 (C4 (Web Text))

Max Activating Token Index: 264

Text #1

Max Range: 4.1274. Min Range: -4.1274

Max Act: 4.0023. Min Act: -0.1700

Data Index: 1258666 (C4 (Web Text))

Max Activating Token Index: 479

Text #2

Max Range: 4.1274. Min Range: -4.1274

Max Act: 4.0023. Min Act: -0.1700

Data Index: 1494125 (Python Code)

Max Activating Token Index: 217

Text #3

Max Range: 4.1274. Min Range: -4.1274

Max Act: 3.9710. Min Act: -0.1700

Data Index: 171876 (C4 (Web Text))

Max Activating Token Index: 582

Text #4

Max Range: 4.1274. Min Range: -4.1274

Max Act: 3.9398. Min Act: -0.1700

Data Index: 1435537 (Python Code)

Max Activating Token Index: 990

Text #5

Max Range: 4.1274. Min Range: -4.1274

Max Act: 3.9398. Min Act: -0.1700

Data Index: 166338 (C4 (Web Text))

Max Activating Token Index: 447

Text #6

Max Range: 4.1274. Min Range: -4.1274

Max Act: 3.9398. Min Act: -0.1700

Data Index: 1158175 (C4 (Web Text))

Max Activating Token Index: 399

Text #7

Max Range: 4.1274. Min Range: -4.1274

Max Act: 3.9085. Min Act: -0.1700

Data Index: 1383110 (Python Code)

Max Activating Token Index: 990

Text #8

Max Range: 4.1274. Min Range: -4.1274

Max Act: 3.9710. Min Act: -0.1700

Data Index: 1629644 (Python Code)

Max Activating Token Index: 74

Text #9

Max Range: 4.1274. Min Range: -4.1274

Max Act: 3.9085. Min Act: -0.1700

Data Index: 1545138 (Python Code)

Max Activating Token Index: 745

Text #10

Max Range: 4.1274. Min Range: -4.1274

Max Act: 3.9398. Min Act: -0.1700

Data Index: 1500815 (Python Code)

Max Activating Token Index: 602

Text #11

Max Range: 4.1274. Min Range: -4.1274

Max Act: 3.9085. Min Act: -0.1700

Data Index: 350382 (C4 (Web Text))

Max Activating Token Index: 735

Text #12

Max Range: 4.1274. Min Range: -4.1274

Max Act: 3.8772. Min Act: -0.1700

Data Index: 725677 (C4 (Web Text))

Max Activating Token Index: 452

Text #13

Max Range: 4.1274. Min Range: -4.1274

Max Act: 3.8772. Min Act: -0.1700

Data Index: 454235 (C4 (Web Text))

Max Activating Token Index: 13

Text #14

Max Range: 4.1274. Min Range: -4.1274

Max Act: 3.8460. Min Act: -0.1700

Data Index: 1489596 (Python Code)

Max Activating Token Index: 91

Text #15

Max Range: 4.1274. Min Range: -4.1274

Max Act: 3.8772. Min Act: -0.1700

Data Index: 1430845 (Python Code)

Max Activating Token Index: 274

Text #16

Max Range: 4.1274. Min Range: -4.1274

Max Act: 3.8460. Min Act: -0.1700

Data Index: 1550077 (Python Code)

Max Activating Token Index: 131

Text #17

Max Range: 4.1274. Min Range: -4.1274

Max Act: 3.9085. Min Act: -0.1700

Data Index: 1533346 (Python Code)

Max Activating Token Index: 1021

Text #18

Max Range: 4.1274. Min Range: -4.1274

Max Act: 3.8460. Min Act: -0.1700

Data Index: 232329 (C4 (Web Text))

Max Activating Token Index: 912

Text #19

Max Range: 4.1274. Min Range: -4.1274

Max Act: 3.8772. Min Act: -0.1700

Data Index: 861269 (C4 (Web Text))

Max Activating Token Index: 800