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

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

Neuron 1450 in Layer 1

Load this data into an Interactive Neuroscope

See Documentation here

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

Text #0

Max Range: 3.6827. Min Range: -3.6827

Max Act: 3.6827. Min Act: -0.1700

Data Index: 1282010 (C4 (Web Text))

Max Activating Token Index: 860

Text #1

Max Range: 3.6827. Min Range: -3.6827

Max Act: 3.5887. Min Act: -0.1700

Data Index: 520880 (C4 (Web Text))

Max Activating Token Index: 922

Text #2

Max Range: 3.6827. Min Range: -3.6827

Max Act: 3.5887. Min Act: -0.1700

Data Index: 837565 (C4 (Web Text))

Max Activating Token Index: 823

Text #3

Max Range: 3.6827. Min Range: -3.6827

Max Act: 3.5574. Min Act: -0.1700

Data Index: 416567 (C4 (Web Text))

Max Activating Token Index: 935

Text #4

Max Range: 3.6827. Min Range: -3.6827

Max Act: 3.5887. Min Act: -0.1700

Data Index: 1277697 (C4 (Web Text))

Max Activating Token Index: 1013

Text #5

Max Range: 3.6827. Min Range: -3.6827

Max Act: 3.4948. Min Act: -0.1700

Data Index: 665747 (C4 (Web Text))

Max Activating Token Index: 323

Text #6

Max Range: 3.6827. Min Range: -3.6827

Max Act: 3.4791. Min Act: -0.1700

Data Index: 625333 (C4 (Web Text))

Max Activating Token Index: 165

Text #7

Max Range: 3.6827. Min Range: -3.6827

Max Act: 3.4948. Min Act: -0.1700

Data Index: 1149631 (C4 (Web Text))

Max Activating Token Index: 8

Text #8

Max Range: 3.6827. Min Range: -3.6827

Max Act: 3.4791. Min Act: -0.1700

Data Index: 573868 (C4 (Web Text))

Max Activating Token Index: 339

Text #9

Max Range: 3.6827. Min Range: -3.6827

Max Act: 3.5261. Min Act: -0.1700

Data Index: 929156 (C4 (Web Text))

Max Activating Token Index: 871

Text #10

Max Range: 3.6827. Min Range: -3.6827

Max Act: 3.4948. Min Act: -0.1700

Data Index: 154004 (C4 (Web Text))

Max Activating Token Index: 372

Text #11

Max Range: 3.6827. Min Range: -3.6827

Max Act: 3.4791. Min Act: -0.1700

Data Index: 503304 (C4 (Web Text))

Max Activating Token Index: 525

Text #12

Max Range: 3.6827. Min Range: -3.6827

Max Act: 3.4164. Min Act: -0.1700

Data Index: 1114754 (C4 (Web Text))

Max Activating Token Index: 638

Text #13

Max Range: 3.6827. Min Range: -3.6827

Max Act: 3.4321. Min Act: -0.1700

Data Index: 1298802 (C4 (Web Text))

Max Activating Token Index: 195

Text #14

Max Range: 3.6827. Min Range: -3.6827

Max Act: 3.4164. Min Act: -0.1700

Data Index: 855660 (C4 (Web Text))

Max Activating Token Index: 99

Text #15

Max Range: 3.6827. Min Range: -3.6827

Max Act: 3.4321. Min Act: -0.1700

Data Index: 210353 (C4 (Web Text))

Max Activating Token Index: 305

Text #16

Max Range: 3.6827. Min Range: -3.6827

Max Act: 3.3850. Min Act: -0.1700

Data Index: 844132 (C4 (Web Text))

Max Activating Token Index: 252

Text #17

Max Range: 3.6827. Min Range: -3.6827

Max Act: 3.3693. Min Act: -0.1700

Data Index: 1107363 (C4 (Web Text))

Max Activating Token Index: 898

Text #18

Max Range: 3.6827. Min Range: -3.6827

Max Act: 3.3850. Min Act: -0.1700

Data Index: 677637 (C4 (Web Text))

Max Activating Token Index: 209

Text #19

Max Range: 3.6827. Min Range: -3.6827

Max Act: 3.3693. Min Act: -0.1700

Data Index: 240370 (C4 (Web Text))

Max Activating Token Index: 900