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

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

Neuron 2040 in Layer 1

Load this data into an Interactive Neuroscope

See Documentation here

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

Text #0

Max Range: 3.8223. Min Range: -3.8223

Max Act: 3.8223. Min Act: -0.1700

Data Index: 373048 (C4 (Web Text))

Max Activating Token Index: 32

Text #1

Max Range: 3.8223. Min Range: -3.8223

Max Act: 3.7910. Min Act: -0.1700

Data Index: 613178 (C4 (Web Text))

Max Activating Token Index: 85

Text #2

Max Range: 3.8223. Min Range: -3.8223

Max Act: 3.7753. Min Act: -0.1700

Data Index: 293222 (C4 (Web Text))

Max Activating Token Index: 201

Text #3

Max Range: 3.8223. Min Range: -3.8223

Max Act: 3.7597. Min Act: -0.1700

Data Index: 1038543 (C4 (Web Text))

Max Activating Token Index: 404

Text #4

Max Range: 3.8223. Min Range: -3.8223

Max Act: 3.7597. Min Act: -0.1700

Data Index: 89760 (C4 (Web Text))

Max Activating Token Index: 981

Text #5

Max Range: 3.8223. Min Range: -3.8223

Max Act: 3.7597. Min Act: -0.1700

Data Index: 96497 (C4 (Web Text))

Max Activating Token Index: 65

Text #6

Max Range: 3.8223. Min Range: -3.8223

Max Act: 3.7597. Min Act: -0.1700

Data Index: 550771 (C4 (Web Text))

Max Activating Token Index: 711

Text #7

Max Range: 3.8223. Min Range: -3.8223

Max Act: 3.7910. Min Act: -0.1700

Data Index: 635171 (C4 (Web Text))

Max Activating Token Index: 769

Text #8

Max Range: 3.8223. Min Range: -3.8223

Max Act: 3.7441. Min Act: -0.1700

Data Index: 1083444 (C4 (Web Text))

Max Activating Token Index: 734

Text #9

Max Range: 3.8223. Min Range: -3.8223

Max Act: 3.7441. Min Act: -0.1700

Data Index: 365385 (C4 (Web Text))

Max Activating Token Index: 734

Text #10

Max Range: 3.8223. Min Range: -3.8223

Max Act: 3.7597. Min Act: -0.1700

Data Index: 676045 (C4 (Web Text))

Max Activating Token Index: 712

Text #11

Max Range: 3.8223. Min Range: -3.8223

Max Act: 3.7753. Min Act: -0.1700

Data Index: 370994 (C4 (Web Text))

Max Activating Token Index: 555

Text #12

Max Range: 3.8223. Min Range: -3.8223

Max Act: 3.7441. Min Act: -0.1700

Data Index: 273450 (C4 (Web Text))

Max Activating Token Index: 974

Text #13

Max Range: 3.8223. Min Range: -3.8223

Max Act: 3.7441. Min Act: -0.1700

Data Index: 212873 (C4 (Web Text))

Max Activating Token Index: 456

Text #14

Max Range: 3.8223. Min Range: -3.8223

Max Act: 3.7284. Min Act: -0.1700

Data Index: 17135 (C4 (Web Text))

Max Activating Token Index: 899

Text #15

Max Range: 3.8223. Min Range: -3.8223

Max Act: 3.7441. Min Act: -0.1700

Data Index: 735023 (C4 (Web Text))

Max Activating Token Index: 511

Text #16

Max Range: 3.8223. Min Range: -3.8223

Max Act: 3.7284. Min Act: -0.1700

Data Index: 739921 (C4 (Web Text))

Max Activating Token Index: 25

Text #17

Max Range: 3.8223. Min Range: -3.8223

Max Act: 3.7753. Min Act: -0.1700

Data Index: 810501 (C4 (Web Text))

Max Activating Token Index: 958

Text #18

Max Range: 3.8223. Min Range: -3.8223

Max Act: 3.7597. Min Act: -0.1700

Data Index: 974283 (C4 (Web Text))

Max Activating Token Index: 429

Text #19

Max Range: 3.8223. Min Range: -3.8223

Max Act: 3.7441. Min Act: -0.1700

Data Index: 938076 (C4 (Web Text))

Max Activating Token Index: 348