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

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

Neuron 1050 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: 5.3214. Min Range: -5.3214

Max Act: 5.3214. Min Act: -0.1700

Data Index: 483735 (C4 (Web Text))

Max Activating Token Index: 816

Text #1

Max Range: 5.3214. Min Range: -5.3214

Max Act: 5.2901. Min Act: -0.1700

Data Index: 571334 (C4 (Web Text))

Max Activating Token Index: 497

Text #2

Max Range: 5.3214. Min Range: -5.3214

Max Act: 5.2901. Min Act: -0.1700

Data Index: 1158190 (C4 (Web Text))

Max Activating Token Index: 808

Text #3

Max Range: 5.3214. Min Range: -5.3214

Max Act: 5.2276. Min Act: -0.1700

Data Index: 711687 (C4 (Web Text))

Max Activating Token Index: 499

Text #4

Max Range: 5.3214. Min Range: -5.3214

Max Act: 5.2276. Min Act: -0.1700

Data Index: 366340 (C4 (Web Text))

Max Activating Token Index: 1002

Text #5

Max Range: 5.3214. Min Range: -5.3214

Max Act: 5.1339. Min Act: -0.1700

Data Index: 500911 (C4 (Web Text))

Max Activating Token Index: 664

Text #6

Max Range: 5.3214. Min Range: -5.3214

Max Act: 5.1026. Min Act: -0.1700

Data Index: 515461 (C4 (Web Text))

Max Activating Token Index: 841

Text #7

Max Range: 5.3214. Min Range: -5.3214

Max Act: 5.1026. Min Act: -0.1700

Data Index: 1340907 (C4 (Web Text))

Max Activating Token Index: 742

Text #8

Max Range: 5.3214. Min Range: -5.3214

Max Act: 5.1339. Min Act: -0.1700

Data Index: 395851 (C4 (Web Text))

Max Activating Token Index: 1018

Text #9

Max Range: 5.3214. Min Range: -5.3214

Max Act: 5.0714. Min Act: -0.1700

Data Index: 535817 (C4 (Web Text))

Max Activating Token Index: 69

Text #10

Max Range: 5.3214. Min Range: -5.3214

Max Act: 5.1339. Min Act: -0.1700

Data Index: 911603 (C4 (Web Text))

Max Activating Token Index: 52

Text #11

Max Range: 5.3214. Min Range: -5.3214

Max Act: 5.0714. Min Act: -0.1700

Data Index: 903372 (C4 (Web Text))

Max Activating Token Index: 249

Text #12

Max Range: 5.3214. Min Range: -5.3214

Max Act: 5.1026. Min Act: -0.1700

Data Index: 1082038 (C4 (Web Text))

Max Activating Token Index: 681

Text #13

Max Range: 5.3214. Min Range: -5.3214

Max Act: 5.0401. Min Act: -0.1700

Data Index: 405603 (C4 (Web Text))

Max Activating Token Index: 167

Text #14

Max Range: 5.3214. Min Range: -5.3214

Max Act: 5.0714. Min Act: -0.1700

Data Index: 974880 (C4 (Web Text))

Max Activating Token Index: 803

Text #15

Max Range: 5.3214. Min Range: -5.3214

Max Act: 5.0714. Min Act: -0.1700

Data Index: 1354669 (C4 (Web Text))

Max Activating Token Index: 581

Text #16

Max Range: 5.3214. Min Range: -5.3214

Max Act: 5.0714. Min Act: -0.1700

Data Index: 1100652 (C4 (Web Text))

Max Activating Token Index: 599

Text #17

Max Range: 5.3214. Min Range: -5.3214

Max Act: 5.0714. Min Act: -0.1700

Data Index: 62937 (C4 (Web Text))

Max Activating Token Index: 988

Text #18

Max Range: 5.3214. Min Range: -5.3214

Max Act: 5.0714. Min Act: -0.1700

Data Index: 1060921 (C4 (Web Text))

Max Activating Token Index: 615

Text #19

Max Range: 5.3214. Min Range: -5.3214

Max Act: 5.0714. Min Act: -0.1700

Data Index: 451653 (C4 (Web Text))

Max Activating Token Index: 213