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

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

Neuron 12 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: 8.3602. Min Range: -8.3602

Max Act: 8.3602. Min Act: -0.1700

Data Index: 143248 (C4 (Web Text))

Max Activating Token Index: 940

Text #1

Max Range: 8.3602. Min Range: -8.3602

Max Act: 8.2352. Min Act: -0.1700

Data Index: 140982 (C4 (Web Text))

Max Activating Token Index: 345

Text #2

Max Range: 8.3602. Min Range: -8.3602

Max Act: 8.2977. Min Act: -0.0702

Data Index: 393196 (C4 (Web Text))

Max Activating Token Index: 417

Text #3

Max Range: 8.3602. Min Range: -8.3602

Max Act: 8.2352. Min Act: -0.1692

Data Index: 181684 (C4 (Web Text))

Max Activating Token Index: 319

Text #4

Max Range: 8.3602. Min Range: -8.3602

Max Act: 8.1102. Min Act: -0.1699

Data Index: 763767 (C4 (Web Text))

Max Activating Token Index: 370

Text #5

Max Range: 8.3602. Min Range: -8.3602

Max Act: 8.1102. Min Act: -0.1699

Data Index: 1335531 (C4 (Web Text))

Max Activating Token Index: 569

Text #6

Max Range: 8.3602. Min Range: -8.3602

Max Act: 8.1727. Min Act: -0.1658

Data Index: 1168750 (C4 (Web Text))

Max Activating Token Index: 250

Text #7

Max Range: 8.3602. Min Range: -8.3602

Max Act: 7.9227. Min Act: -0.1700

Data Index: 986571 (C4 (Web Text))

Max Activating Token Index: 703

Text #8

Max Range: 8.3602. Min Range: -8.3602

Max Act: 7.9852. Min Act: -0.1441

Data Index: 111898 (C4 (Web Text))

Max Activating Token Index: 519

Text #9

Max Range: 8.3602. Min Range: -8.3602

Max Act: 7.9852. Min Act: -0.1700

Data Index: 152956 (C4 (Web Text))

Max Activating Token Index: 917

Text #10

Max Range: 8.3602. Min Range: -8.3602

Max Act: 7.9227. Min Act: -0.1695

Data Index: 1118956 (C4 (Web Text))

Max Activating Token Index: 796

Text #11

Max Range: 8.3602. Min Range: -8.3602

Max Act: 7.9227. Min Act: -0.1700

Data Index: 43665 (C4 (Web Text))

Max Activating Token Index: 306

Text #12

Max Range: 8.3602. Min Range: -8.3602

Max Act: 7.9227. Min Act: -0.1700

Data Index: 123763 (C4 (Web Text))

Max Activating Token Index: 36

Text #13

Max Range: 8.3602. Min Range: -8.3602

Max Act: 7.8602. Min Act: -0.1700

Data Index: 1204614 (C4 (Web Text))

Max Activating Token Index: 632

Text #14

Max Range: 8.3602. Min Range: -8.3602

Max Act: 7.8602. Min Act: -0.1699

Data Index: 890296 (C4 (Web Text))

Max Activating Token Index: 272

Text #15

Max Range: 8.3602. Min Range: -8.3602

Max Act: 7.8602. Min Act: -0.1681

Data Index: 1123806 (C4 (Web Text))

Max Activating Token Index: 827

Text #16

Max Range: 8.3602. Min Range: -8.3602

Max Act: 7.9227. Min Act: -0.1698

Data Index: 29574 (C4 (Web Text))

Max Activating Token Index: 80

Text #17

Max Range: 8.3602. Min Range: -8.3602

Max Act: 7.8602. Min Act: -0.1700

Data Index: 524713 (C4 (Web Text))

Max Activating Token Index: 735

Text #18

Max Range: 8.3602. Min Range: -8.3602

Max Act: 7.9227. Min Act: -0.1700

Data Index: 1021388 (C4 (Web Text))

Max Activating Token Index: 447

Text #19

Max Range: 8.3602. Min Range: -8.3602

Max Act: 7.8602. Min Act: -0.1700

Data Index: 558975 (C4 (Web Text))

Max Activating Token Index: 256