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

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

Neuron 1195 in Layer 2

Load this data into an Interactive Neuroscope

See Documentation here

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

Text #0

Max Range: 7.1079. Min Range: -7.1079

Max Act: 7.1079. Min Act: -0.1699

Data Index: 612816 (C4 (Web Text))

Max Activating Token Index: 863

Text #1

Max Range: 7.1079. Min Range: -7.1079

Max Act: 7.0454. Min Act: -0.1699

Data Index: 1267525 (C4 (Web Text))

Max Activating Token Index: 426

Text #2

Max Range: 7.1079. Min Range: -7.1079

Max Act: 6.8892. Min Act: -0.1700

Data Index: 233328 (C4 (Web Text))

Max Activating Token Index: 961

Text #3

Max Range: 7.1079. Min Range: -7.1079

Max Act: 6.7954. Min Act: -0.1699

Data Index: 587456 (C4 (Web Text))

Max Activating Token Index: 327

Text #4

Max Range: 7.1079. Min Range: -7.1079

Max Act: 6.8267. Min Act: -0.1700

Data Index: 594599 (C4 (Web Text))

Max Activating Token Index: 478

Text #5

Max Range: 7.1079. Min Range: -7.1079

Max Act: 6.8579. Min Act: -0.1699

Data Index: 733800 (C4 (Web Text))

Max Activating Token Index: 980

Text #6

Max Range: 7.1079. Min Range: -7.1079

Max Act: 6.8579. Min Act: -0.1700

Data Index: 1357999 (C4 (Web Text))

Max Activating Token Index: 853

Text #7

Max Range: 7.1079. Min Range: -7.1079

Max Act: 6.6704. Min Act: -0.1700

Data Index: 903068 (C4 (Web Text))

Max Activating Token Index: 182

Text #8

Max Range: 7.1079. Min Range: -7.1079

Max Act: 6.7017. Min Act: -0.1700

Data Index: 697845 (C4 (Web Text))

Max Activating Token Index: 218

Text #9

Max Range: 7.1079. Min Range: -7.1079

Max Act: 6.7642. Min Act: -0.1700

Data Index: 1338197 (C4 (Web Text))

Max Activating Token Index: 748

Text #10

Max Range: 7.1079. Min Range: -7.1079

Max Act: 6.6392. Min Act: -0.1700

Data Index: 363713 (C4 (Web Text))

Max Activating Token Index: 736

Text #11

Max Range: 7.1079. Min Range: -7.1079

Max Act: 6.7329. Min Act: -0.1700

Data Index: 324586 (C4 (Web Text))

Max Activating Token Index: 124

Text #12

Max Range: 7.1079. Min Range: -7.1079

Max Act: 6.7954. Min Act: -0.1699

Data Index: 400476 (C4 (Web Text))

Max Activating Token Index: 655

Text #13

Max Range: 7.1079. Min Range: -7.1079

Max Act: 6.7017. Min Act: -0.1699

Data Index: 1276462 (C4 (Web Text))

Max Activating Token Index: 315

Text #14

Max Range: 7.1079. Min Range: -7.1079

Max Act: 6.6392. Min Act: -0.1699

Data Index: 798322 (C4 (Web Text))

Max Activating Token Index: 218

Text #15

Max Range: 7.1079. Min Range: -7.1079

Max Act: 6.7642. Min Act: -0.1700

Data Index: 795026 (C4 (Web Text))

Max Activating Token Index: 137

Text #16

Max Range: 7.1079. Min Range: -7.1079

Max Act: 6.6079. Min Act: -0.1700

Data Index: 1184283 (C4 (Web Text))

Max Activating Token Index: 978

Text #17

Max Range: 7.1079. Min Range: -7.1079

Max Act: 6.5767. Min Act: -0.1699

Data Index: 39112 (C4 (Web Text))

Max Activating Token Index: 422

Text #18

Max Range: 7.1079. Min Range: -7.1079

Max Act: 6.5767. Min Act: -0.1700

Data Index: 1007365 (C4 (Web Text))

Max Activating Token Index: 728

Text #19

Max Range: 7.1079. Min Range: -7.1079

Max Act: 6.6392. Min Act: -0.1700

Data Index: 142529 (C4 (Web Text))

Max Activating Token Index: 564