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

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

Neuron 1028 in Layer 1

Load this data into an Interactive Neuroscope

See Documentation here

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

Text #0

Max Range: 3.2080. Min Range: -3.2080

Max Act: 3.2080. Min Act: -0.1700

Data Index: 939847 (C4 (Web Text))

Max Activating Token Index: 672

Text #1

Max Range: 3.2080. Min Range: -3.2080

Max Act: 3.1923. Min Act: -0.1700

Data Index: 586426 (C4 (Web Text))

Max Activating Token Index: 804

Text #2

Max Range: 3.2080. Min Range: -3.2080

Max Act: 3.0978. Min Act: -0.1700

Data Index: 1197835 (C4 (Web Text))

Max Activating Token Index: 255

Text #3

Max Range: 3.2080. Min Range: -3.2080

Max Act: 3.0820. Min Act: -0.1700

Data Index: 1358527 (C4 (Web Text))

Max Activating Token Index: 481

Text #4

Max Range: 3.2080. Min Range: -3.2080

Max Act: 3.0505. Min Act: -0.1700

Data Index: 744616 (C4 (Web Text))

Max Activating Token Index: 557

Text #5

Max Range: 3.2080. Min Range: -3.2080

Max Act: 3.0347. Min Act: -0.1700

Data Index: 917214 (C4 (Web Text))

Max Activating Token Index: 984

Text #6

Max Range: 3.2080. Min Range: -3.2080

Max Act: 3.0347. Min Act: -0.1700

Data Index: 942880 (C4 (Web Text))

Max Activating Token Index: 825

Text #7

Max Range: 3.2080. Min Range: -3.2080

Max Act: 3.0031. Min Act: -0.1700

Data Index: 436436 (C4 (Web Text))

Max Activating Token Index: 823

Text #8

Max Range: 3.2080. Min Range: -3.2080

Max Act: 3.0189. Min Act: -0.1700

Data Index: 1040362 (C4 (Web Text))

Max Activating Token Index: 345

Text #9

Max Range: 3.2080. Min Range: -3.2080

Max Act: 3.0189. Min Act: -0.1700

Data Index: 483335 (C4 (Web Text))

Max Activating Token Index: 611

Text #10

Max Range: 3.2080. Min Range: -3.2080

Max Act: 2.9873. Min Act: -0.1700

Data Index: 317922 (C4 (Web Text))

Max Activating Token Index: 64

Text #11

Max Range: 3.2080. Min Range: -3.2080

Max Act: 2.9873. Min Act: -0.1700

Data Index: 657361 (C4 (Web Text))

Max Activating Token Index: 755

Text #12

Max Range: 3.2080. Min Range: -3.2080

Max Act: 2.9714. Min Act: -0.1700

Data Index: 1359017 (C4 (Web Text))

Max Activating Token Index: 326

Text #13

Max Range: 3.2080. Min Range: -3.2080

Max Act: 3.0031. Min Act: -0.1700

Data Index: 155625 (C4 (Web Text))

Max Activating Token Index: 645

Text #14

Max Range: 3.2080. Min Range: -3.2080

Max Act: 2.9873. Min Act: -0.1700

Data Index: 305116 (C4 (Web Text))

Max Activating Token Index: 15

Text #15

Max Range: 3.2080. Min Range: -3.2080

Max Act: 2.9714. Min Act: -0.1700

Data Index: 1648921 (Python Code)

Max Activating Token Index: 19

Text #16

Max Range: 3.2080. Min Range: -3.2080

Max Act: 2.9556. Min Act: -0.1700

Data Index: 175314 (C4 (Web Text))

Max Activating Token Index: 138

Text #17

Max Range: 3.2080. Min Range: -3.2080

Max Act: 3.0031. Min Act: -0.1700

Data Index: 547057 (C4 (Web Text))

Max Activating Token Index: 656

Text #18

Max Range: 3.2080. Min Range: -3.2080

Max Act: 2.9714. Min Act: -0.1700

Data Index: 1141009 (C4 (Web Text))

Max Activating Token Index: 83

Text #19

Max Range: 3.2080. Min Range: -3.2080

Max Act: 2.9398. Min Act: -0.1700

Data Index: 521588 (C4 (Web Text))

Max Activating Token Index: 556