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

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

Neuron 675 in Layer 3

Load this data into an Interactive Neuroscope

See Documentation here

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

Text #0

Max Range: 5.2955. Min Range: -5.2955

Max Act: 5.2955. Min Act: -0.1700

Data Index: 1061810 (C4 (Web Text))

Max Activating Token Index: 896

Text #1

Max Range: 5.2955. Min Range: -5.2955

Max Act: 5.1393. Min Act: -0.1700

Data Index: 1147527 (C4 (Web Text))

Max Activating Token Index: 181

Text #2

Max Range: 5.2955. Min Range: -5.2955

Max Act: 5.1080. Min Act: -0.1700

Data Index: 272459 (C4 (Web Text))

Max Activating Token Index: 169

Text #3

Max Range: 5.2955. Min Range: -5.2955

Max Act: 5.0768. Min Act: -0.1700

Data Index: 375711 (C4 (Web Text))

Max Activating Token Index: 974

Text #4

Max Range: 5.2955. Min Range: -5.2955

Max Act: 4.7955. Min Act: -0.1700

Data Index: 380371 (C4 (Web Text))

Max Activating Token Index: 461

Text #5

Max Range: 5.2955. Min Range: -5.2955

Max Act: 4.7174. Min Act: -0.1700

Data Index: 149985 (C4 (Web Text))

Max Activating Token Index: 997

Text #6

Max Range: 5.2955. Min Range: -5.2955

Max Act: 4.7643. Min Act: -0.1700

Data Index: 407894 (C4 (Web Text))

Max Activating Token Index: 870

Text #7

Max Range: 5.2955. Min Range: -5.2955

Max Act: 4.7330. Min Act: -0.1700

Data Index: 711495 (C4 (Web Text))

Max Activating Token Index: 883

Text #8

Max Range: 5.2955. Min Range: -5.2955

Max Act: 4.7018. Min Act: -0.1700

Data Index: 1285149 (C4 (Web Text))

Max Activating Token Index: 906

Text #9

Max Range: 5.2955. Min Range: -5.2955

Max Act: 4.6861. Min Act: -0.1700

Data Index: 819617 (C4 (Web Text))

Max Activating Token Index: 912

Text #10

Max Range: 5.2955. Min Range: -5.2955

Max Act: 4.6080. Min Act: -0.1700

Data Index: 1075866 (C4 (Web Text))

Max Activating Token Index: 553

Text #11

Max Range: 5.2955. Min Range: -5.2955

Max Act: 4.6080. Min Act: -0.1700

Data Index: 1123120 (C4 (Web Text))

Max Activating Token Index: 761

Text #12

Max Range: 5.2955. Min Range: -5.2955

Max Act: 4.6705. Min Act: -0.1700

Data Index: 1220365 (C4 (Web Text))

Max Activating Token Index: 793

Text #13

Max Range: 5.2955. Min Range: -5.2955

Max Act: 4.5924. Min Act: -0.1700

Data Index: 1489867 (Python Code)

Max Activating Token Index: 558

Text #14

Max Range: 5.2955. Min Range: -5.2955

Max Act: 4.5611. Min Act: -0.1700

Data Index: 497258 (C4 (Web Text))

Max Activating Token Index: 754

Text #15

Max Range: 5.2955. Min Range: -5.2955

Max Act: 4.5768. Min Act: -0.1700

Data Index: 485701 (C4 (Web Text))

Max Activating Token Index: 932

Text #16

Max Range: 5.2955. Min Range: -5.2955

Max Act: 4.5611. Min Act: -0.1700

Data Index: 1105686 (C4 (Web Text))

Max Activating Token Index: 743

Text #17

Max Range: 5.2955. Min Range: -5.2955

Max Act: 4.5455. Min Act: -0.1700

Data Index: 1244408 (C4 (Web Text))

Max Activating Token Index: 966

Text #18

Max Range: 5.2955. Min Range: -5.2955

Max Act: 4.5455. Min Act: -0.1700

Data Index: 1341893 (C4 (Web Text))

Max Activating Token Index: 617

Text #19

Max Range: 5.2955. Min Range: -5.2955

Max Act: 4.5611. Min Act: -0.1700

Data Index: 760129 (C4 (Web Text))

Max Activating Token Index: 985