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

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

Neuron 50 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: 5.3539. Min Range: -5.3539

Max Act: 5.3226. Min Act: -0.1699

Data Index: 295648 (C4 (Web Text))

Max Activating Token Index: 815

Text #1

Max Range: 5.3539. Min Range: -5.3539

Max Act: 5.3539. Min Act: -0.1700

Data Index: 1117260 (C4 (Web Text))

Max Activating Token Index: 918

Text #2

Max Range: 5.3539. Min Range: -5.3539

Max Act: 5.1976. Min Act: -0.1700

Data Index: 1204350 (C4 (Web Text))

Max Activating Token Index: 865

Text #3

Max Range: 5.3539. Min Range: -5.3539

Max Act: 5.2914. Min Act: -0.1700

Data Index: 948031 (C4 (Web Text))

Max Activating Token Index: 940

Text #4

Max Range: 5.3539. Min Range: -5.3539

Max Act: 5.2601. Min Act: -0.1700

Data Index: 847317 (C4 (Web Text))

Max Activating Token Index: 752

Text #5

Max Range: 5.3539. Min Range: -5.3539

Max Act: 5.2601. Min Act: -0.1700

Data Index: 59259 (C4 (Web Text))

Max Activating Token Index: 989

Text #6

Max Range: 5.3539. Min Range: -5.3539

Max Act: 5.1664. Min Act: -0.1699

Data Index: 688187 (C4 (Web Text))

Max Activating Token Index: 936

Text #7

Max Range: 5.3539. Min Range: -5.3539

Max Act: 5.1039. Min Act: -0.1700

Data Index: 228081 (C4 (Web Text))

Max Activating Token Index: 783

Text #8

Max Range: 5.3539. Min Range: -5.3539

Max Act: 5.3226. Min Act: -0.1699

Data Index: 479581 (C4 (Web Text))

Max Activating Token Index: 781

Text #9

Max Range: 5.3539. Min Range: -5.3539

Max Act: 5.1664. Min Act: -0.1700

Data Index: 628423 (C4 (Web Text))

Max Activating Token Index: 821

Text #10

Max Range: 5.3539. Min Range: -5.3539

Max Act: 5.2289. Min Act: -0.1700

Data Index: 1154873 (C4 (Web Text))

Max Activating Token Index: 870

Text #11

Max Range: 5.3539. Min Range: -5.3539

Max Act: 5.2289. Min Act: -0.1700

Data Index: 73593 (C4 (Web Text))

Max Activating Token Index: 990

Text #12

Max Range: 5.3539. Min Range: -5.3539

Max Act: 5.1664. Min Act: -0.1700

Data Index: 585884 (C4 (Web Text))

Max Activating Token Index: 791

Text #13

Max Range: 5.3539. Min Range: -5.3539

Max Act: 5.1664. Min Act: -0.1700

Data Index: 1252314 (C4 (Web Text))

Max Activating Token Index: 812

Text #14

Max Range: 5.3539. Min Range: -5.3539

Max Act: 5.1664. Min Act: -0.1700

Data Index: 456693 (C4 (Web Text))

Max Activating Token Index: 865

Text #15

Max Range: 5.3539. Min Range: -5.3539

Max Act: 5.1976. Min Act: -0.1700

Data Index: 1134127 (C4 (Web Text))

Max Activating Token Index: 819

Text #16

Max Range: 5.3539. Min Range: -5.3539

Max Act: 5.2289. Min Act: -0.1700

Data Index: 1136313 (C4 (Web Text))

Max Activating Token Index: 906

Text #17

Max Range: 5.3539. Min Range: -5.3539

Max Act: 5.1976. Min Act: -0.1700

Data Index: 1160231 (C4 (Web Text))

Max Activating Token Index: 910

Text #18

Max Range: 5.3539. Min Range: -5.3539

Max Act: 5.2289. Min Act: -0.1700

Data Index: 501698 (C4 (Web Text))

Max Activating Token Index: 170

Text #19

Max Range: 5.3539. Min Range: -5.3539

Max Act: 5.1664. Min Act: -0.1700

Data Index: 202610 (C4 (Web Text))

Max Activating Token Index: 888