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

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

Neuron 731 in Layer 2

Load this data into an Interactive Neuroscope

See Documentation here

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

Text #0

Max Range: 5.5657. Min Range: -5.5657

Max Act: 5.5032. Min Act: -0.1700

Data Index: 377196 (C4 (Web Text))

Max Activating Token Index: 680

Text #1

Max Range: 5.5657. Min Range: -5.5657

Max Act: 5.5657. Min Act: -0.1700

Data Index: 62341 (C4 (Web Text))

Max Activating Token Index: 790

Text #2

Max Range: 5.5657. Min Range: -5.5657

Max Act: 5.3157. Min Act: -0.1700

Data Index: 1407117 (Python Code)

Max Activating Token Index: 561

Text #3

Max Range: 5.5657. Min Range: -5.5657

Max Act: 5.2532. Min Act: -0.1700

Data Index: 1245538 (C4 (Web Text))

Max Activating Token Index: 743

Text #4

Max Range: 5.5657. Min Range: -5.5657

Max Act: 5.1594. Min Act: -0.1700

Data Index: 1174533 (C4 (Web Text))

Max Activating Token Index: 876

Text #5

Max Range: 5.5657. Min Range: -5.5657

Max Act: 5.2844. Min Act: -0.1700

Data Index: 1046854 (C4 (Web Text))

Max Activating Token Index: 494

Text #6

Max Range: 5.5657. Min Range: -5.5657

Max Act: 5.3157. Min Act: -0.1700

Data Index: 1427041 (Python Code)

Max Activating Token Index: 118

Text #7

Max Range: 5.5657. Min Range: -5.5657

Max Act: 5.1282. Min Act: -0.1700

Data Index: 1113785 (C4 (Web Text))

Max Activating Token Index: 965

Text #8

Max Range: 5.5657. Min Range: -5.5657

Max Act: 5.2844. Min Act: -0.1700

Data Index: 1576639 (Python Code)

Max Activating Token Index: 399

Text #9

Max Range: 5.5657. Min Range: -5.5657

Max Act: 5.3157. Min Act: -0.1700

Data Index: 332905 (C4 (Web Text))

Max Activating Token Index: 854

Text #10

Max Range: 5.5657. Min Range: -5.5657

Max Act: 5.1282. Min Act: -0.1700

Data Index: 1055199 (C4 (Web Text))

Max Activating Token Index: 966

Text #11

Max Range: 5.5657. Min Range: -5.5657

Max Act: 5.0657. Min Act: -0.1700

Data Index: 1020272 (C4 (Web Text))

Max Activating Token Index: 514

Text #12

Max Range: 5.5657. Min Range: -5.5657

Max Act: 5.1282. Min Act: -0.1700

Data Index: 1590208 (Python Code)

Max Activating Token Index: 329

Text #13

Max Range: 5.5657. Min Range: -5.5657

Max Act: 5.1907. Min Act: -0.1700

Data Index: 1170469 (C4 (Web Text))

Max Activating Token Index: 657

Text #14

Max Range: 5.5657. Min Range: -5.5657

Max Act: 5.2844. Min Act: -0.1700

Data Index: 344418 (C4 (Web Text))

Max Activating Token Index: 659

Text #15

Max Range: 5.5657. Min Range: -5.5657

Max Act: 5.1907. Min Act: -0.1699

Data Index: 292456 (C4 (Web Text))

Max Activating Token Index: 247

Text #16

Max Range: 5.5657. Min Range: -5.5657

Max Act: 5.1907. Min Act: -0.1700

Data Index: 542531 (C4 (Web Text))

Max Activating Token Index: 767

Text #17

Max Range: 5.5657. Min Range: -5.5657

Max Act: 5.1594. Min Act: -0.1700

Data Index: 1018005 (C4 (Web Text))

Max Activating Token Index: 565

Text #18

Max Range: 5.5657. Min Range: -5.5657

Max Act: 4.9719. Min Act: -0.1700

Data Index: 35425 (C4 (Web Text))

Max Activating Token Index: 378

Text #19

Max Range: 5.5657. Min Range: -5.5657

Max Act: 5.0032. Min Act: -0.1700

Data Index: 647359 (C4 (Web Text))

Max Activating Token Index: 922