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

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

Neuron 1232 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: 3.7849. Min Range: -3.7849

Max Act: 3.7536. Min Act: -0.1700

Data Index: 498877 (C4 (Web Text))

Max Activating Token Index: 534

Text #1

Max Range: 3.7849. Min Range: -3.7849

Max Act: 3.7849. Min Act: -0.1700

Data Index: 1548416 (Python Code)

Max Activating Token Index: 161

Text #2

Max Range: 3.7849. Min Range: -3.7849

Max Act: 3.7536. Min Act: -0.1700

Data Index: 1417637 (Python Code)

Max Activating Token Index: 78

Text #3

Max Range: 3.7849. Min Range: -3.7849

Max Act: 3.7223. Min Act: -0.1700

Data Index: 1498022 (Python Code)

Max Activating Token Index: 367

Text #4

Max Range: 3.7849. Min Range: -3.7849

Max Act: 3.7536. Min Act: -0.1700

Data Index: 1521442 (Python Code)

Max Activating Token Index: 796

Text #5

Max Range: 3.7849. Min Range: -3.7849

Max Act: 3.7536. Min Act: -0.1700

Data Index: 1483853 (Python Code)

Max Activating Token Index: 598

Text #6

Max Range: 3.7849. Min Range: -3.7849

Max Act: 3.6910. Min Act: -0.1700

Data Index: 695619 (C4 (Web Text))

Max Activating Token Index: 83

Text #7

Max Range: 3.7849. Min Range: -3.7849

Max Act: 3.6910. Min Act: -0.1700

Data Index: 1558101 (Python Code)

Max Activating Token Index: 862

Text #8

Max Range: 3.7849. Min Range: -3.7849

Max Act: 3.6597. Min Act: -0.1700

Data Index: 9932 (C4 (Web Text))

Max Activating Token Index: 973

Text #9

Max Range: 3.7849. Min Range: -3.7849

Max Act: 3.6597. Min Act: -0.1700

Data Index: 172316 (C4 (Web Text))

Max Activating Token Index: 715

Text #10

Max Range: 3.7849. Min Range: -3.7849

Max Act: 3.6284. Min Act: -0.1700

Data Index: 1072809 (C4 (Web Text))

Max Activating Token Index: 691

Text #11

Max Range: 3.7849. Min Range: -3.7849

Max Act: 3.6597. Min Act: -0.1700

Data Index: 1637004 (Python Code)

Max Activating Token Index: 766

Text #12

Max Range: 3.7849. Min Range: -3.7849

Max Act: 3.6597. Min Act: -0.1700

Data Index: 138726 (C4 (Web Text))

Max Activating Token Index: 956

Text #13

Max Range: 3.7849. Min Range: -3.7849

Max Act: 3.6597. Min Act: -0.1700

Data Index: 823030 (C4 (Web Text))

Max Activating Token Index: 648

Text #14

Max Range: 3.7849. Min Range: -3.7849

Max Act: 3.6597. Min Act: -0.1700

Data Index: 663284 (C4 (Web Text))

Max Activating Token Index: 774

Text #15

Max Range: 3.7849. Min Range: -3.7849

Max Act: 3.6597. Min Act: -0.1700

Data Index: 118174 (C4 (Web Text))

Max Activating Token Index: 838

Text #16

Max Range: 3.7849. Min Range: -3.7849

Max Act: 3.5971. Min Act: -0.1700

Data Index: 413715 (C4 (Web Text))

Max Activating Token Index: 970

Text #17

Max Range: 3.7849. Min Range: -3.7849

Max Act: 3.6284. Min Act: -0.1700

Data Index: 1490087 (Python Code)

Max Activating Token Index: 299

Text #18

Max Range: 3.7849. Min Range: -3.7849

Max Act: 3.6284. Min Act: -0.1700

Data Index: 1622721 (Python Code)

Max Activating Token Index: 540

Text #19

Max Range: 3.7849. Min Range: -3.7849

Max Act: 3.6284. Min Act: -0.1700

Data Index: 847735 (C4 (Web Text))

Max Activating Token Index: 448