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

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

Neuron 163 in Layer 2

Load this data into an Interactive Neuroscope

See Documentation here

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

Text #0

Max Range: 5.9549. Min Range: -5.9549

Max Act: 5.9549. Min Act: -0.1700

Data Index: 516110 (C4 (Web Text))

Max Activating Token Index: 658

Text #1

Max Range: 5.9549. Min Range: -5.9549

Max Act: 5.7362. Min Act: -0.1700

Data Index: 54642 (C4 (Web Text))

Max Activating Token Index: 842

Text #2

Max Range: 5.9549. Min Range: -5.9549

Max Act: 5.4862. Min Act: -0.1700

Data Index: 736864 (C4 (Web Text))

Max Activating Token Index: 512

Text #3

Max Range: 5.9549. Min Range: -5.9549

Max Act: 5.5799. Min Act: -0.1700

Data Index: 212979 (C4 (Web Text))

Max Activating Token Index: 246

Text #4

Max Range: 5.9549. Min Range: -5.9549

Max Act: 5.4862. Min Act: -0.1700

Data Index: 1605738 (Python Code)

Max Activating Token Index: 410

Text #5

Max Range: 5.9549. Min Range: -5.9549

Max Act: 5.4549. Min Act: -0.1700

Data Index: 32101 (C4 (Web Text))

Max Activating Token Index: 221

Text #6

Max Range: 5.9549. Min Range: -5.9549

Max Act: 5.4549. Min Act: -0.1700

Data Index: 118823 (C4 (Web Text))

Max Activating Token Index: 759

Text #7

Max Range: 5.9549. Min Range: -5.9549

Max Act: 5.3924. Min Act: -0.1700

Data Index: 1022892 (C4 (Web Text))

Max Activating Token Index: 725

Text #8

Max Range: 5.9549. Min Range: -5.9549

Max Act: 5.4237. Min Act: -0.1700

Data Index: 860083 (C4 (Web Text))

Max Activating Token Index: 311

Text #9

Max Range: 5.9549. Min Range: -5.9549

Max Act: 5.4549. Min Act: -0.1700

Data Index: 772828 (C4 (Web Text))

Max Activating Token Index: 446

Text #10

Max Range: 5.9549. Min Range: -5.9549

Max Act: 5.4549. Min Act: -0.1700

Data Index: 359156 (C4 (Web Text))

Max Activating Token Index: 815

Text #11

Max Range: 5.9549. Min Range: -5.9549

Max Act: 5.3612. Min Act: -0.1700

Data Index: 917850 (C4 (Web Text))

Max Activating Token Index: 614

Text #12

Max Range: 5.9549. Min Range: -5.9549

Max Act: 5.3924. Min Act: -0.1700

Data Index: 41881 (C4 (Web Text))

Max Activating Token Index: 930

Text #13

Max Range: 5.9549. Min Range: -5.9549

Max Act: 5.2674. Min Act: -0.1700

Data Index: 1059722 (C4 (Web Text))

Max Activating Token Index: 753

Text #14

Max Range: 5.9549. Min Range: -5.9549

Max Act: 5.2987. Min Act: -0.1700

Data Index: 1368292 (Python Code)

Max Activating Token Index: 809

Text #15

Max Range: 5.9549. Min Range: -5.9549

Max Act: 5.3299. Min Act: -0.1700

Data Index: 887446 (C4 (Web Text))

Max Activating Token Index: 963

Text #16

Max Range: 5.9549. Min Range: -5.9549

Max Act: 5.2674. Min Act: -0.1700

Data Index: 193144 (C4 (Web Text))

Max Activating Token Index: 695

Text #17

Max Range: 5.9549. Min Range: -5.9549

Max Act: 5.1737. Min Act: -0.1700

Data Index: 608 (C4 (Web Text))

Max Activating Token Index: 460

Text #18

Max Range: 5.9549. Min Range: -5.9549

Max Act: 5.2674. Min Act: -0.1700

Data Index: 586415 (C4 (Web Text))

Max Activating Token Index: 261

Text #19

Max Range: 5.9549. Min Range: -5.9549

Max Act: 5.3612. Min Act: -0.1700

Data Index: 183611 (C4 (Web Text))

Max Activating Token Index: 849