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

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

Neuron 1459 in Layer 3

Load this data into an Interactive Neuroscope

See Documentation here

Transformer Lens Loading: HookedTransformer.from_pretrained('gelu-4l')

Text #0

Max Range: 3.7678. Min Range: -3.7678

Max Act: 3.7678. Min Act: -0.1700

Data Index: 407372 (C4 (Web Text))

Max Activating Token Index: 965

Text #1

Max Range: 3.7678. Min Range: -3.7678

Max Act: 3.6427. Min Act: -0.1700

Data Index: 388396 (C4 (Web Text))

Max Activating Token Index: 334

Text #2

Max Range: 3.7678. Min Range: -3.7678

Max Act: 3.6427. Min Act: -0.1700

Data Index: 1158657 (C4 (Web Text))

Max Activating Token Index: 352

Text #3

Max Range: 3.7678. Min Range: -3.7678

Max Act: 3.5957. Min Act: -0.1700

Data Index: 284464 (C4 (Web Text))

Max Activating Token Index: 680

Text #4

Max Range: 3.7678. Min Range: -3.7678

Max Act: 3.6270. Min Act: -0.1700

Data Index: 1135803 (C4 (Web Text))

Max Activating Token Index: 680

Text #5

Max Range: 3.7678. Min Range: -3.7678

Max Act: 3.5800. Min Act: -0.1700

Data Index: 1037539 (C4 (Web Text))

Max Activating Token Index: 952

Text #6

Max Range: 3.7678. Min Range: -3.7678

Max Act: 3.5487. Min Act: -0.1700

Data Index: 38510 (C4 (Web Text))

Max Activating Token Index: 648

Text #7

Max Range: 3.7678. Min Range: -3.7678

Max Act: 3.5487. Min Act: -0.1700

Data Index: 575322 (C4 (Web Text))

Max Activating Token Index: 663

Text #8

Max Range: 3.7678. Min Range: -3.7678

Max Act: 3.5644. Min Act: -0.1700

Data Index: 373065 (C4 (Web Text))

Max Activating Token Index: 203

Text #9

Max Range: 3.7678. Min Range: -3.7678

Max Act: 3.5487. Min Act: -0.1700

Data Index: 796834 (C4 (Web Text))

Max Activating Token Index: 116

Text #10

Max Range: 3.7678. Min Range: -3.7678

Max Act: 3.5487. Min Act: -0.1700

Data Index: 1279383 (C4 (Web Text))

Max Activating Token Index: 855

Text #11

Max Range: 3.7678. Min Range: -3.7678

Max Act: 3.5487. Min Act: -0.1700

Data Index: 1029974 (C4 (Web Text))

Max Activating Token Index: 765

Text #12

Max Range: 3.7678. Min Range: -3.7678

Max Act: 3.5330. Min Act: -0.1700

Data Index: 610968 (C4 (Web Text))

Max Activating Token Index: 534

Text #13

Max Range: 3.7678. Min Range: -3.7678

Max Act: 3.5174. Min Act: -0.1700

Data Index: 502948 (C4 (Web Text))

Max Activating Token Index: 159

Text #14

Max Range: 3.7678. Min Range: -3.7678

Max Act: 3.5017. Min Act: -0.1700

Data Index: 1419707 (Python Code)

Max Activating Token Index: 894

Text #15

Max Range: 3.7678. Min Range: -3.7678

Max Act: 3.5017. Min Act: -0.1700

Data Index: 1072072 (C4 (Web Text))

Max Activating Token Index: 577

Text #16

Max Range: 3.7678. Min Range: -3.7678

Max Act: 3.4860. Min Act: -0.1700

Data Index: 176454 (C4 (Web Text))

Max Activating Token Index: 375

Text #17

Max Range: 3.7678. Min Range: -3.7678

Max Act: 3.5017. Min Act: -0.1700

Data Index: 254951 (C4 (Web Text))

Max Activating Token Index: 844

Text #18

Max Range: 3.7678. Min Range: -3.7678

Max Act: 3.5017. Min Act: -0.1700

Data Index: 1306600 (C4 (Web Text))

Max Activating Token Index: 533

Text #19

Max Range: 3.7678. Min Range: -3.7678

Max Act: 3.5174. Min Act: -0.1700

Data Index: 541079 (C4 (Web Text))

Max Activating Token Index: 923