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

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

Neuron 1392 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: 3.4926. Min Range: -3.4926

Max Act: 3.4926. Min Act: -0.1700

Data Index: 703486 (C4 (Web Text))

Max Activating Token Index: 57

Text #1

Max Range: 3.4926. Min Range: -3.4926

Max Act: 3.4299. Min Act: -0.1700

Data Index: 34738 (C4 (Web Text))

Max Activating Token Index: 110

Text #2

Max Range: 3.4926. Min Range: -3.4926

Max Act: 3.4299. Min Act: -0.1700

Data Index: 1481840 (Python Code)

Max Activating Token Index: 509

Text #3

Max Range: 3.4926. Min Range: -3.4926

Max Act: 3.4299. Min Act: -0.1700

Data Index: 1072741 (C4 (Web Text))

Max Activating Token Index: 824

Text #4

Max Range: 3.4926. Min Range: -3.4926

Max Act: 3.4299. Min Act: -0.1700

Data Index: 245766 (C4 (Web Text))

Max Activating Token Index: 683

Text #5

Max Range: 3.4926. Min Range: -3.4926

Max Act: 3.4299. Min Act: -0.1700

Data Index: 434125 (C4 (Web Text))

Max Activating Token Index: 652

Text #6

Max Range: 3.4926. Min Range: -3.4926

Max Act: 3.4613. Min Act: -0.1700

Data Index: 345291 (C4 (Web Text))

Max Activating Token Index: 579

Text #7

Max Range: 3.4926. Min Range: -3.4926

Max Act: 3.3672. Min Act: -0.1700

Data Index: 494189 (C4 (Web Text))

Max Activating Token Index: 663

Text #8

Max Range: 3.4926. Min Range: -3.4926

Max Act: 3.3672. Min Act: -0.1700

Data Index: 755164 (C4 (Web Text))

Max Activating Token Index: 858

Text #9

Max Range: 3.4926. Min Range: -3.4926

Max Act: 3.4299. Min Act: -0.1700

Data Index: 900558 (C4 (Web Text))

Max Activating Token Index: 494

Text #10

Max Range: 3.4926. Min Range: -3.4926

Max Act: 3.3672. Min Act: -0.1700

Data Index: 1492319 (Python Code)

Max Activating Token Index: 622

Text #11

Max Range: 3.4926. Min Range: -3.4926

Max Act: 3.3358. Min Act: -0.1700

Data Index: 1521893 (Python Code)

Max Activating Token Index: 72

Text #12

Max Range: 3.4926. Min Range: -3.4926

Max Act: 3.3672. Min Act: -0.1700

Data Index: 789896 (C4 (Web Text))

Max Activating Token Index: 495

Text #13

Max Range: 3.4926. Min Range: -3.4926

Max Act: 3.3358. Min Act: -0.1700

Data Index: 226794 (C4 (Web Text))

Max Activating Token Index: 536

Text #14

Max Range: 3.4926. Min Range: -3.4926

Max Act: 3.3986. Min Act: -0.1700

Data Index: 538594 (C4 (Web Text))

Max Activating Token Index: 887

Text #15

Max Range: 3.4926. Min Range: -3.4926

Max Act: 3.3358. Min Act: -0.1700

Data Index: 1547574 (Python Code)

Max Activating Token Index: 547

Text #16

Max Range: 3.4926. Min Range: -3.4926

Max Act: 3.3044. Min Act: -0.1700

Data Index: 1564996 (Python Code)

Max Activating Token Index: 87

Text #17

Max Range: 3.4926. Min Range: -3.4926

Max Act: 3.3044. Min Act: -0.1700

Data Index: 1570287 (Python Code)

Max Activating Token Index: 748

Text #18

Max Range: 3.4926. Min Range: -3.4926

Max Act: 3.3044. Min Act: -0.1700

Data Index: 1531144 (Python Code)

Max Activating Token Index: 887

Text #19

Max Range: 3.4926. Min Range: -3.4926

Max Act: 3.3358. Min Act: -0.1700

Data Index: 1450640 (Python Code)

Max Activating Token Index: 881