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

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

Neuron 591 in Layer 0

Load this data into an Interactive Neuroscope

See Documentation here

Transformer Lens Loading: HookedTransformer.from_pretrained('gelu-3l')

Text #0

Max Range: 3.3569. Min Range: -3.3569

Max Act: 3.3569. Min Act: -0.1700

Data Index: 1254021 (C4 (Web Text))

Max Activating Token Index: 452

Text #1

Max Range: 3.3569. Min Range: -3.3569

Max Act: 3.3412. Min Act: -0.1700

Data Index: 63791 (C4 (Web Text))

Max Activating Token Index: 628

Text #2

Max Range: 3.3569. Min Range: -3.3569

Max Act: 3.3255. Min Act: -0.1700

Data Index: 1019072 (C4 (Web Text))

Max Activating Token Index: 864

Text #3

Max Range: 3.3569. Min Range: -3.3569

Max Act: 3.3098. Min Act: -0.1700

Data Index: 163846 (C4 (Web Text))

Max Activating Token Index: 978

Text #4

Max Range: 3.3569. Min Range: -3.3569

Max Act: 3.3098. Min Act: -0.1700

Data Index: 201801 (C4 (Web Text))

Max Activating Token Index: 518

Text #5

Max Range: 3.3569. Min Range: -3.3569

Max Act: 3.2941. Min Act: -0.1700

Data Index: 692995 (C4 (Web Text))

Max Activating Token Index: 313

Text #6

Max Range: 3.3569. Min Range: -3.3569

Max Act: 3.1841. Min Act: -0.1700

Data Index: 545015 (C4 (Web Text))

Max Activating Token Index: 884

Text #7

Max Range: 3.3569. Min Range: -3.3569

Max Act: 3.1841. Min Act: -0.1700

Data Index: 1351602 (C4 (Web Text))

Max Activating Token Index: 273

Text #8

Max Range: 3.3569. Min Range: -3.3569

Max Act: 3.1683. Min Act: -0.1700

Data Index: 193038 (C4 (Web Text))

Max Activating Token Index: 36

Text #9

Max Range: 3.3569. Min Range: -3.3569

Max Act: 3.1526. Min Act: -0.1700

Data Index: 1001961 (C4 (Web Text))

Max Activating Token Index: 112

Text #10

Max Range: 3.3569. Min Range: -3.3569

Max Act: 3.1526. Min Act: -0.1700

Data Index: 530449 (C4 (Web Text))

Max Activating Token Index: 875

Text #11

Max Range: 3.3569. Min Range: -3.3569

Max Act: 3.1368. Min Act: -0.1700

Data Index: 427900 (C4 (Web Text))

Max Activating Token Index: 357

Text #12

Max Range: 3.3569. Min Range: -3.3569

Max Act: 3.1368. Min Act: -0.1700

Data Index: 137832 (C4 (Web Text))

Max Activating Token Index: 609

Text #13

Max Range: 3.3569. Min Range: -3.3569

Max Act: 3.1211. Min Act: -0.1700

Data Index: 47234 (C4 (Web Text))

Max Activating Token Index: 918

Text #14

Max Range: 3.3569. Min Range: -3.3569

Max Act: 3.0896. Min Act: -0.1700

Data Index: 1229690 (C4 (Web Text))

Max Activating Token Index: 103

Text #15

Max Range: 3.3569. Min Range: -3.3569

Max Act: 3.1053. Min Act: -0.1700

Data Index: 1204612 (C4 (Web Text))

Max Activating Token Index: 24

Text #16

Max Range: 3.3569. Min Range: -3.3569

Max Act: 3.0896. Min Act: -0.1700

Data Index: 1493076 (Python Code)

Max Activating Token Index: 2

Text #17

Max Range: 3.3569. Min Range: -3.3569

Max Act: 3.0738. Min Act: -0.1700

Data Index: 805739 (C4 (Web Text))

Max Activating Token Index: 890

Text #18

Max Range: 3.3569. Min Range: -3.3569

Max Act: 3.0580. Min Act: -0.1700

Data Index: 1043687 (C4 (Web Text))

Max Activating Token Index: 358

Text #19

Max Range: 3.3569. Min Range: -3.3569

Max Act: 3.0580. Min Act: -0.1700

Data Index: 866560 (C4 (Web Text))

Max Activating Token Index: 66