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

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

Neuron 546 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: 5.8617. Min Range: -5.8617

Max Act: 5.8617. Min Act: -0.1700

Data Index: 93910 (C4 (Web Text))

Max Activating Token Index: 850

Text #1

Max Range: 5.8617. Min Range: -5.8617

Max Act: 5.7367. Min Act: -0.1700

Data Index: 896930 (C4 (Web Text))

Max Activating Token Index: 310

Text #2

Max Range: 5.8617. Min Range: -5.8617

Max Act: 5.5805. Min Act: -0.1700

Data Index: 968197 (C4 (Web Text))

Max Activating Token Index: 546

Text #3

Max Range: 5.8617. Min Range: -5.8617

Max Act: 5.5492. Min Act: -0.1700

Data Index: 1134799 (C4 (Web Text))

Max Activating Token Index: 842

Text #4

Max Range: 5.8617. Min Range: -5.8617

Max Act: 5.5492. Min Act: -0.1700

Data Index: 1139400 (C4 (Web Text))

Max Activating Token Index: 784

Text #5

Max Range: 5.8617. Min Range: -5.8617

Max Act: 5.5180. Min Act: -0.1700

Data Index: 487075 (C4 (Web Text))

Max Activating Token Index: 988

Text #6

Max Range: 5.8617. Min Range: -5.8617

Max Act: 5.4867. Min Act: -0.1700

Data Index: 396801 (C4 (Web Text))

Max Activating Token Index: 743

Text #7

Max Range: 5.8617. Min Range: -5.8617

Max Act: 5.4555. Min Act: -0.1700

Data Index: 266101 (C4 (Web Text))

Max Activating Token Index: 633

Text #8

Max Range: 5.8617. Min Range: -5.8617

Max Act: 5.4555. Min Act: -0.1700

Data Index: 549443 (C4 (Web Text))

Max Activating Token Index: 364

Text #9

Max Range: 5.8617. Min Range: -5.8617

Max Act: 5.4555. Min Act: -0.1700

Data Index: 943352 (C4 (Web Text))

Max Activating Token Index: 84

Text #10

Max Range: 5.8617. Min Range: -5.8617

Max Act: 5.4555. Min Act: -0.1700

Data Index: 653657 (C4 (Web Text))

Max Activating Token Index: 681

Text #11

Max Range: 5.8617. Min Range: -5.8617

Max Act: 5.4555. Min Act: -0.1700

Data Index: 499425 (C4 (Web Text))

Max Activating Token Index: 478

Text #12

Max Range: 5.8617. Min Range: -5.8617

Max Act: 5.4242. Min Act: -0.1700

Data Index: 1296090 (C4 (Web Text))

Max Activating Token Index: 198

Text #13

Max Range: 5.8617. Min Range: -5.8617

Max Act: 5.4242. Min Act: -0.1700

Data Index: 886855 (C4 (Web Text))

Max Activating Token Index: 857

Text #14

Max Range: 5.8617. Min Range: -5.8617

Max Act: 5.4555. Min Act: -0.1700

Data Index: 108030 (C4 (Web Text))

Max Activating Token Index: 166

Text #15

Max Range: 5.8617. Min Range: -5.8617

Max Act: 5.4242. Min Act: -0.1699

Data Index: 342693 (C4 (Web Text))

Max Activating Token Index: 641

Text #16

Max Range: 5.8617. Min Range: -5.8617

Max Act: 5.4867. Min Act: -0.1700

Data Index: 799600 (C4 (Web Text))

Max Activating Token Index: 771

Text #17

Max Range: 5.8617. Min Range: -5.8617

Max Act: 5.4242. Min Act: -0.1700

Data Index: 1219419 (C4 (Web Text))

Max Activating Token Index: 776

Text #18

Max Range: 5.8617. Min Range: -5.8617

Max Act: 5.3930. Min Act: -0.1700

Data Index: 639069 (C4 (Web Text))

Max Activating Token Index: 743

Text #19

Max Range: 5.8617. Min Range: -5.8617

Max Act: 5.3617. Min Act: -0.1700

Data Index: 297835 (C4 (Web Text))

Max Activating Token Index: 916