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

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

Neuron 1673 in Layer 1

Load this data into an Interactive Neuroscope

See Documentation here

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

Text #0

Max Range: 3.5964. Min Range: -3.5964

Max Act: 3.5964. Min Act: -0.1700

Data Index: 1307193 (C4 (Web Text))

Max Activating Token Index: 542

Text #1

Max Range: 3.5964. Min Range: -3.5964

Max Act: 3.4867. Min Act: -0.1700

Data Index: 1246719 (C4 (Web Text))

Max Activating Token Index: 607

Text #2

Max Range: 3.5964. Min Range: -3.5964

Max Act: 3.4397. Min Act: -0.1700

Data Index: 592019 (C4 (Web Text))

Max Activating Token Index: 659

Text #3

Max Range: 3.5964. Min Range: -3.5964

Max Act: 3.4240. Min Act: -0.1700

Data Index: 1245293 (C4 (Web Text))

Max Activating Token Index: 525

Text #4

Max Range: 3.5964. Min Range: -3.5964

Max Act: 3.3613. Min Act: -0.1700

Data Index: 755730 (C4 (Web Text))

Max Activating Token Index: 837

Text #5

Max Range: 3.5964. Min Range: -3.5964

Max Act: 3.3927. Min Act: -0.1700

Data Index: 913304 (C4 (Web Text))

Max Activating Token Index: 21

Text #6

Max Range: 3.5964. Min Range: -3.5964

Max Act: 3.4397. Min Act: -0.1700

Data Index: 353397 (C4 (Web Text))

Max Activating Token Index: 447

Text #7

Max Range: 3.5964. Min Range: -3.5964

Max Act: 3.4240. Min Act: -0.1700

Data Index: 1347518 (C4 (Web Text))

Max Activating Token Index: 744

Text #8

Max Range: 3.5964. Min Range: -3.5964

Max Act: 3.4397. Min Act: -0.1700

Data Index: 489150 (C4 (Web Text))

Max Activating Token Index: 649

Text #9

Max Range: 3.5964. Min Range: -3.5964

Max Act: 3.3770. Min Act: -0.1700

Data Index: 315790 (C4 (Web Text))

Max Activating Token Index: 811

Text #10

Max Range: 3.5964. Min Range: -3.5964

Max Act: 3.4083. Min Act: -0.1700

Data Index: 112136 (C4 (Web Text))

Max Activating Token Index: 191

Text #11

Max Range: 3.5964. Min Range: -3.5964

Max Act: 3.3770. Min Act: -0.1700

Data Index: 1064098 (C4 (Web Text))

Max Activating Token Index: 867

Text #12

Max Range: 3.5964. Min Range: -3.5964

Max Act: 3.3927. Min Act: -0.1700

Data Index: 1151601 (C4 (Web Text))

Max Activating Token Index: 944

Text #13

Max Range: 3.5964. Min Range: -3.5964

Max Act: 3.3770. Min Act: -0.1700

Data Index: 352035 (C4 (Web Text))

Max Activating Token Index: 262

Text #14

Max Range: 3.5964. Min Range: -3.5964

Max Act: 3.3613. Min Act: -0.1700

Data Index: 998688 (C4 (Web Text))

Max Activating Token Index: 696

Text #15

Max Range: 3.5964. Min Range: -3.5964

Max Act: 3.3613. Min Act: -0.1700

Data Index: 543357 (C4 (Web Text))

Max Activating Token Index: 382

Text #16

Max Range: 3.5964. Min Range: -3.5964

Max Act: 3.4083. Min Act: -0.1700

Data Index: 962661 (C4 (Web Text))

Max Activating Token Index: 865

Text #17

Max Range: 3.5964. Min Range: -3.5964

Max Act: 3.2985. Min Act: -0.1700

Data Index: 200800 (C4 (Web Text))

Max Activating Token Index: 786

Text #18

Max Range: 3.5964. Min Range: -3.5964

Max Act: 3.3456. Min Act: -0.1700

Data Index: 882625 (C4 (Web Text))

Max Activating Token Index: 748

Text #19

Max Range: 3.5964. Min Range: -3.5964

Max Act: 3.3299. Min Act: -0.1700

Data Index: 1097471 (C4 (Web Text))

Max Activating Token Index: 322