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

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

Neuron 390 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: 5.1722. Min Range: -5.1722

Max Act: 5.1722. Min Act: -0.1700

Data Index: 609910 (C4 (Web Text))

Max Activating Token Index: 626

Text #1

Max Range: 5.1722. Min Range: -5.1722

Max Act: 5.1410. Min Act: -0.1700

Data Index: 387756 (C4 (Web Text))

Max Activating Token Index: 584

Text #2

Max Range: 5.1722. Min Range: -5.1722

Max Act: 5.1097. Min Act: -0.1700

Data Index: 1181110 (C4 (Web Text))

Max Activating Token Index: 109

Text #3

Max Range: 5.1722. Min Range: -5.1722

Max Act: 5.0785. Min Act: -0.1700

Data Index: 1226299 (C4 (Web Text))

Max Activating Token Index: 335

Text #4

Max Range: 5.1722. Min Range: -5.1722

Max Act: 5.0160. Min Act: -0.1700

Data Index: 957480 (C4 (Web Text))

Max Activating Token Index: 578

Text #5

Max Range: 5.1722. Min Range: -5.1722

Max Act: 4.9535. Min Act: -0.1700

Data Index: 877895 (C4 (Web Text))

Max Activating Token Index: 628

Text #6

Max Range: 5.1722. Min Range: -5.1722

Max Act: 4.9222. Min Act: -0.1700

Data Index: 868055 (C4 (Web Text))

Max Activating Token Index: 547

Text #7

Max Range: 5.1722. Min Range: -5.1722

Max Act: 4.8597. Min Act: -0.1700

Data Index: 1323830 (C4 (Web Text))

Max Activating Token Index: 500

Text #8

Max Range: 5.1722. Min Range: -5.1722

Max Act: 4.9535. Min Act: -0.1700

Data Index: 1257154 (C4 (Web Text))

Max Activating Token Index: 752

Text #9

Max Range: 5.1722. Min Range: -5.1722

Max Act: 4.8910. Min Act: -0.1700

Data Index: 32699 (C4 (Web Text))

Max Activating Token Index: 454

Text #10

Max Range: 5.1722. Min Range: -5.1722

Max Act: 4.8597. Min Act: -0.1700

Data Index: 1063762 (C4 (Web Text))

Max Activating Token Index: 394

Text #11

Max Range: 5.1722. Min Range: -5.1722

Max Act: 4.8910. Min Act: -0.1700

Data Index: 933922 (C4 (Web Text))

Max Activating Token Index: 72

Text #12

Max Range: 5.1722. Min Range: -5.1722

Max Act: 4.8597. Min Act: -0.1700

Data Index: 1151211 (C4 (Web Text))

Max Activating Token Index: 878

Text #13

Max Range: 5.1722. Min Range: -5.1722

Max Act: 4.8597. Min Act: -0.1700

Data Index: 804513 (C4 (Web Text))

Max Activating Token Index: 923

Text #14

Max Range: 5.1722. Min Range: -5.1722

Max Act: 4.8597. Min Act: -0.1699

Data Index: 1212807 (C4 (Web Text))

Max Activating Token Index: 460

Text #15

Max Range: 5.1722. Min Range: -5.1722

Max Act: 4.7972. Min Act: -0.1700

Data Index: 510457 (C4 (Web Text))

Max Activating Token Index: 573

Text #16

Max Range: 5.1722. Min Range: -5.1722

Max Act: 4.7972. Min Act: -0.1700

Data Index: 758801 (C4 (Web Text))

Max Activating Token Index: 792

Text #17

Max Range: 5.1722. Min Range: -5.1722

Max Act: 4.7972. Min Act: -0.1700

Data Index: 521688 (C4 (Web Text))

Max Activating Token Index: 386

Text #18

Max Range: 5.1722. Min Range: -5.1722

Max Act: 4.8285. Min Act: -0.1700

Data Index: 459898 (C4 (Web Text))

Max Activating Token Index: 121

Text #19

Max Range: 5.1722. Min Range: -5.1722

Max Act: 4.7972. Min Act: -0.1700

Data Index: 619841 (C4 (Web Text))

Max Activating Token Index: 855