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

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

Neuron 138 in Layer 1

Load this data into an Interactive Neuroscope

See Documentation here

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

Text #0

Max Range: 3.9563. Min Range: -3.9563

Max Act: 3.9563. Min Act: -0.1700

Data Index: 975758 (C4 (Web Text))

Max Activating Token Index: 355

Text #1

Max Range: 3.9563. Min Range: -3.9563

Max Act: 3.8938. Min Act: -0.1700

Data Index: 1116033 (C4 (Web Text))

Max Activating Token Index: 858

Text #2

Max Range: 3.9563. Min Range: -3.9563

Max Act: 3.8938. Min Act: -0.1700

Data Index: 1399905 (Python Code)

Max Activating Token Index: 771

Text #3

Max Range: 3.9563. Min Range: -3.9563

Max Act: 3.8625. Min Act: -0.1700

Data Index: 1546031 (Python Code)

Max Activating Token Index: 707

Text #4

Max Range: 3.9563. Min Range: -3.9563

Max Act: 3.7374. Min Act: -0.1700

Data Index: 174831 (C4 (Web Text))

Max Activating Token Index: 543

Text #5

Max Range: 3.9563. Min Range: -3.9563

Max Act: 3.7999. Min Act: -0.1700

Data Index: 478622 (C4 (Web Text))

Max Activating Token Index: 1018

Text #6

Max Range: 3.9563. Min Range: -3.9563

Max Act: 3.7374. Min Act: -0.1700

Data Index: 258343 (C4 (Web Text))

Max Activating Token Index: 303

Text #7

Max Range: 3.9563. Min Range: -3.9563

Max Act: 3.7374. Min Act: -0.1700

Data Index: 1433477 (Python Code)

Max Activating Token Index: 815

Text #8

Max Range: 3.9563. Min Range: -3.9563

Max Act: 3.7374. Min Act: -0.1700

Data Index: 1604397 (Python Code)

Max Activating Token Index: 718

Text #9

Max Range: 3.9563. Min Range: -3.9563

Max Act: 3.7061. Min Act: -0.1700

Data Index: 1411348 (Python Code)

Max Activating Token Index: 718

Text #10

Max Range: 3.9563. Min Range: -3.9563

Max Act: 3.6435. Min Act: -0.1700

Data Index: 791911 (C4 (Web Text))

Max Activating Token Index: 693

Text #11

Max Range: 3.9563. Min Range: -3.9563

Max Act: 3.6748. Min Act: -0.1700

Data Index: 331492 (C4 (Web Text))

Max Activating Token Index: 424

Text #12

Max Range: 3.9563. Min Range: -3.9563

Max Act: 3.6435. Min Act: -0.1700

Data Index: 446924 (C4 (Web Text))

Max Activating Token Index: 506

Text #13

Max Range: 3.9563. Min Range: -3.9563

Max Act: 3.6748. Min Act: -0.1700

Data Index: 1113176 (C4 (Web Text))

Max Activating Token Index: 366

Text #14

Max Range: 3.9563. Min Range: -3.9563

Max Act: 3.6748. Min Act: -0.1700

Data Index: 1543142 (Python Code)

Max Activating Token Index: 358

Text #15

Max Range: 3.9563. Min Range: -3.9563

Max Act: 3.6435. Min Act: -0.1700

Data Index: 1357846 (C4 (Web Text))

Max Activating Token Index: 987

Text #16

Max Range: 3.9563. Min Range: -3.9563

Max Act: 3.6122. Min Act: -0.1700

Data Index: 985275 (C4 (Web Text))

Max Activating Token Index: 588

Text #17

Max Range: 3.9563. Min Range: -3.9563

Max Act: 3.6122. Min Act: -0.1700

Data Index: 464181 (C4 (Web Text))

Max Activating Token Index: 878

Text #18

Max Range: 3.9563. Min Range: -3.9563

Max Act: 3.6122. Min Act: -0.1700

Data Index: 1477642 (Python Code)

Max Activating Token Index: 399

Text #19

Max Range: 3.9563. Min Range: -3.9563

Max Act: 3.6122. Min Act: -0.1700

Data Index: 1094152 (C4 (Web Text))

Max Activating Token Index: 818