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

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

Neuron 1981 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: 5.0368. Min Range: -5.0368

Max Act: 5.0368. Min Act: -0.1700

Data Index: 560212 (C4 (Web Text))

Max Activating Token Index: 287

Text #1

Max Range: 5.0368. Min Range: -5.0368

Max Act: 4.9743. Min Act: -0.1700

Data Index: 1128227 (C4 (Web Text))

Max Activating Token Index: 152

Text #2

Max Range: 5.0368. Min Range: -5.0368

Max Act: 5.0055. Min Act: -0.1700

Data Index: 1246172 (C4 (Web Text))

Max Activating Token Index: 242

Text #3

Max Range: 5.0368. Min Range: -5.0368

Max Act: 4.9118. Min Act: -0.1700

Data Index: 319 (C4 (Web Text))

Max Activating Token Index: 177

Text #4

Max Range: 5.0368. Min Range: -5.0368

Max Act: 4.8805. Min Act: -0.1700

Data Index: 1255633 (C4 (Web Text))

Max Activating Token Index: 72

Text #5

Max Range: 5.0368. Min Range: -5.0368

Max Act: 4.8493. Min Act: -0.1700

Data Index: 129593 (C4 (Web Text))

Max Activating Token Index: 820

Text #6

Max Range: 5.0368. Min Range: -5.0368

Max Act: 4.8180. Min Act: -0.1700

Data Index: 78724 (C4 (Web Text))

Max Activating Token Index: 729

Text #7

Max Range: 5.0368. Min Range: -5.0368

Max Act: 4.8493. Min Act: -0.1700

Data Index: 499415 (C4 (Web Text))

Max Activating Token Index: 132

Text #8

Max Range: 5.0368. Min Range: -5.0368

Max Act: 4.7868. Min Act: -0.1700

Data Index: 1010278 (C4 (Web Text))

Max Activating Token Index: 145

Text #9

Max Range: 5.0368. Min Range: -5.0368

Max Act: 4.8493. Min Act: -0.1700

Data Index: 515186 (C4 (Web Text))

Max Activating Token Index: 61

Text #10

Max Range: 5.0368. Min Range: -5.0368

Max Act: 4.7868. Min Act: -0.1700

Data Index: 95341 (C4 (Web Text))

Max Activating Token Index: 507

Text #11

Max Range: 5.0368. Min Range: -5.0368

Max Act: 4.8180. Min Act: -0.1700

Data Index: 219320 (C4 (Web Text))

Max Activating Token Index: 290

Text #12

Max Range: 5.0368. Min Range: -5.0368

Max Act: 4.8180. Min Act: -0.1700

Data Index: 962486 (C4 (Web Text))

Max Activating Token Index: 272

Text #13

Max Range: 5.0368. Min Range: -5.0368

Max Act: 4.7243. Min Act: -0.1700

Data Index: 485310 (C4 (Web Text))

Max Activating Token Index: 434

Text #14

Max Range: 5.0368. Min Range: -5.0368

Max Act: 4.7243. Min Act: -0.1700

Data Index: 909657 (C4 (Web Text))

Max Activating Token Index: 378

Text #15

Max Range: 5.0368. Min Range: -5.0368

Max Act: 4.6930. Min Act: -0.1700

Data Index: 1041805 (C4 (Web Text))

Max Activating Token Index: 233

Text #16

Max Range: 5.0368. Min Range: -5.0368

Max Act: 4.6305. Min Act: -0.1700

Data Index: 414039 (C4 (Web Text))

Max Activating Token Index: 83

Text #17

Max Range: 5.0368. Min Range: -5.0368

Max Act: 4.6618. Min Act: -0.1700

Data Index: 327434 (C4 (Web Text))

Max Activating Token Index: 846

Text #18

Max Range: 5.0368. Min Range: -5.0368

Max Act: 4.6930. Min Act: -0.1700

Data Index: 115484 (C4 (Web Text))

Max Activating Token Index: 102

Text #19

Max Range: 5.0368. Min Range: -5.0368

Max Act: 4.6305. Min Act: -0.1700

Data Index: 867402 (C4 (Web Text))

Max Activating Token Index: 647