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

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

Neuron 1333 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: 5.1563. Min Range: -5.1563

Max Act: 5.1563. Min Act: -0.1700

Data Index: 1237677 (C4 (Web Text))

Max Activating Token Index: 197

Text #1

Max Range: 5.1563. Min Range: -5.1563

Max Act: 5.1563. Min Act: -0.1700

Data Index: 1151233 (C4 (Web Text))

Max Activating Token Index: 91

Text #2

Max Range: 5.1563. Min Range: -5.1563

Max Act: 5.1250. Min Act: -0.1700

Data Index: 356518 (C4 (Web Text))

Max Activating Token Index: 527

Text #3

Max Range: 5.1563. Min Range: -5.1563

Max Act: 5.0313. Min Act: -0.1700

Data Index: 1305496 (C4 (Web Text))

Max Activating Token Index: 185

Text #4

Max Range: 5.1563. Min Range: -5.1563

Max Act: 5.0000. Min Act: -0.1700

Data Index: 1053595 (C4 (Web Text))

Max Activating Token Index: 339

Text #5

Max Range: 5.1563. Min Range: -5.1563

Max Act: 4.9688. Min Act: -0.1700

Data Index: 1356261 (C4 (Web Text))

Max Activating Token Index: 93

Text #6

Max Range: 5.1563. Min Range: -5.1563

Max Act: 5.0313. Min Act: -0.1700

Data Index: 384128 (C4 (Web Text))

Max Activating Token Index: 55

Text #7

Max Range: 5.1563. Min Range: -5.1563

Max Act: 4.9688. Min Act: -0.1700

Data Index: 924763 (C4 (Web Text))

Max Activating Token Index: 735

Text #8

Max Range: 5.1563. Min Range: -5.1563

Max Act: 4.9688. Min Act: -0.1700

Data Index: 64319 (C4 (Web Text))

Max Activating Token Index: 336

Text #9

Max Range: 5.1563. Min Range: -5.1563

Max Act: 5.0000. Min Act: -0.1700

Data Index: 688919 (C4 (Web Text))

Max Activating Token Index: 447

Text #10

Max Range: 5.1563. Min Range: -5.1563

Max Act: 4.9375. Min Act: -0.1700

Data Index: 1412155 (Python Code)

Max Activating Token Index: 282

Text #11

Max Range: 5.1563. Min Range: -5.1563

Max Act: 4.9063. Min Act: -0.1700

Data Index: 1307691 (C4 (Web Text))

Max Activating Token Index: 895

Text #12

Max Range: 5.1563. Min Range: -5.1563

Max Act: 4.8750. Min Act: -0.1700

Data Index: 611063 (C4 (Web Text))

Max Activating Token Index: 730

Text #13

Max Range: 5.1563. Min Range: -5.1563

Max Act: 4.8750. Min Act: -0.1700

Data Index: 1323776 (C4 (Web Text))

Max Activating Token Index: 163

Text #14

Max Range: 5.1563. Min Range: -5.1563

Max Act: 4.9063. Min Act: -0.1700

Data Index: 9720 (C4 (Web Text))

Max Activating Token Index: 100

Text #15

Max Range: 5.1563. Min Range: -5.1563

Max Act: 4.8438. Min Act: -0.1700

Data Index: 1289582 (C4 (Web Text))

Max Activating Token Index: 388

Text #16

Max Range: 5.1563. Min Range: -5.1563

Max Act: 4.8750. Min Act: -0.1700

Data Index: 154510 (C4 (Web Text))

Max Activating Token Index: 648

Text #17

Max Range: 5.1563. Min Range: -5.1563

Max Act: 4.9063. Min Act: -0.1700

Data Index: 723719 (C4 (Web Text))

Max Activating Token Index: 132

Text #18

Max Range: 5.1563. Min Range: -5.1563

Max Act: 4.9063. Min Act: -0.1700

Data Index: 727282 (C4 (Web Text))

Max Activating Token Index: 338

Text #19

Max Range: 5.1563. Min Range: -5.1563

Max Act: 4.8750. Min Act: -0.1700

Data Index: 1468826 (Python Code)

Max Activating Token Index: 29