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

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

Neuron 105 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: 5.1868. Min Range: -5.1868

Max Act: 5.1868. Min Act: -0.1700

Data Index: 966191 (C4 (Web Text))

Max Activating Token Index: 759

Text #1

Max Range: 5.1868. Min Range: -5.1868

Max Act: 5.0306. Min Act: -0.1698

Data Index: 524614 (C4 (Web Text))

Max Activating Token Index: 159

Text #2

Max Range: 5.1868. Min Range: -5.1868

Max Act: 5.0306. Min Act: -0.1681

Data Index: 1068976 (C4 (Web Text))

Max Activating Token Index: 201

Text #3

Max Range: 5.1868. Min Range: -5.1868

Max Act: 4.9993. Min Act: -0.1700

Data Index: 966422 (C4 (Web Text))

Max Activating Token Index: 482

Text #4

Max Range: 5.1868. Min Range: -5.1868

Max Act: 4.9993. Min Act: -0.1699

Data Index: 581624 (C4 (Web Text))

Max Activating Token Index: 143

Text #5

Max Range: 5.1868. Min Range: -5.1868

Max Act: 4.9056. Min Act: -0.1700

Data Index: 1356635 (C4 (Web Text))

Max Activating Token Index: 804

Text #6

Max Range: 5.1868. Min Range: -5.1868

Max Act: 4.8118. Min Act: -0.1700

Data Index: 1088298 (C4 (Web Text))

Max Activating Token Index: 914

Text #7

Max Range: 5.1868. Min Range: -5.1868

Max Act: 4.8118. Min Act: -0.1700

Data Index: 744935 (C4 (Web Text))

Max Activating Token Index: 969

Text #8

Max Range: 5.1868. Min Range: -5.1868

Max Act: 4.8118. Min Act: -0.1700

Data Index: 234759 (C4 (Web Text))

Max Activating Token Index: 825

Text #9

Max Range: 5.1868. Min Range: -5.1868

Max Act: 4.8118. Min Act: -0.1700

Data Index: 427437 (C4 (Web Text))

Max Activating Token Index: 454

Text #10

Max Range: 5.1868. Min Range: -5.1868

Max Act: 4.8118. Min Act: -0.1700

Data Index: 1178773 (C4 (Web Text))

Max Activating Token Index: 751

Text #11

Max Range: 5.1868. Min Range: -5.1868

Max Act: 4.7806. Min Act: -0.1700

Data Index: 1292836 (C4 (Web Text))

Max Activating Token Index: 39

Text #12

Max Range: 5.1868. Min Range: -5.1868

Max Act: 4.7806. Min Act: -0.1700

Data Index: 204974 (C4 (Web Text))

Max Activating Token Index: 208

Text #13

Max Range: 5.1868. Min Range: -5.1868

Max Act: 4.7181. Min Act: -0.1700

Data Index: 488861 (C4 (Web Text))

Max Activating Token Index: 983

Text #14

Max Range: 5.1868. Min Range: -5.1868

Max Act: 4.6868. Min Act: -0.1700

Data Index: 587401 (C4 (Web Text))

Max Activating Token Index: 650

Text #15

Max Range: 5.1868. Min Range: -5.1868

Max Act: 4.7806. Min Act: -0.1700

Data Index: 1009244 (C4 (Web Text))

Max Activating Token Index: 613

Text #16

Max Range: 5.1868. Min Range: -5.1868

Max Act: 4.6868. Min Act: -0.1700

Data Index: 685875 (C4 (Web Text))

Max Activating Token Index: 176

Text #17

Max Range: 5.1868. Min Range: -5.1868

Max Act: 4.6556. Min Act: -0.1700

Data Index: 1216060 (C4 (Web Text))

Max Activating Token Index: 669

Text #18

Max Range: 5.1868. Min Range: -5.1868

Max Act: 4.6868. Min Act: -0.1700

Data Index: 353904 (C4 (Web Text))

Max Activating Token Index: 75

Text #19

Max Range: 5.1868. Min Range: -5.1868

Max Act: 4.6868. Min Act: -0.1700

Data Index: 1022028 (C4 (Web Text))

Max Activating Token Index: 817