Model: GELU Model: 2 Layers, 2048 Neurons per Layer

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

Neuron 1616 in Layer 1

Load this data into an Interactive Neuroscope

See Documentation here

Transformer Lens Loading: HookedTransformer.from_pretrained('gelu-2l')

Text #0

Max Range: 3.3468. Min Range: -3.3468

Max Act: 3.3468. Min Act: -0.1700

Data Index: 1317614 (C4 (Web Text))

Max Activating Token Index: 327

Text #1

Max Range: 3.3468. Min Range: -3.3468

Max Act: 3.1897. Min Act: -0.1700

Data Index: 166888 (C4 (Web Text))

Max Activating Token Index: 222

Text #2

Max Range: 3.3468. Min Range: -3.3468

Max Act: 3.1897. Min Act: -0.1700

Data Index: 1482787 (Python Code)

Max Activating Token Index: 106

Text #3

Max Range: 3.3468. Min Range: -3.3468

Max Act: 3.2212. Min Act: -0.1700

Data Index: 595530 (C4 (Web Text))

Max Activating Token Index: 64

Text #4

Max Range: 3.3468. Min Range: -3.3468

Max Act: 3.2212. Min Act: -0.1700

Data Index: 384004 (C4 (Web Text))

Max Activating Token Index: 572

Text #5

Max Range: 3.3468. Min Range: -3.3468

Max Act: 3.2526. Min Act: -0.1700

Data Index: 861193 (C4 (Web Text))

Max Activating Token Index: 467

Text #6

Max Range: 3.3468. Min Range: -3.3468

Max Act: 3.2212. Min Act: -0.1700

Data Index: 354865 (C4 (Web Text))

Max Activating Token Index: 270

Text #7

Max Range: 3.3468. Min Range: -3.3468

Max Act: 3.1897. Min Act: -0.1700

Data Index: 942353 (C4 (Web Text))

Max Activating Token Index: 769

Text #8

Max Range: 3.3468. Min Range: -3.3468

Max Act: 3.1897. Min Act: -0.1700

Data Index: 1309143 (C4 (Web Text))

Max Activating Token Index: 812

Text #9

Max Range: 3.3468. Min Range: -3.3468

Max Act: 3.1267. Min Act: -0.1700

Data Index: 289692 (C4 (Web Text))

Max Activating Token Index: 1023

Text #10

Max Range: 3.3468. Min Range: -3.3468

Max Act: 3.1582. Min Act: -0.1700

Data Index: 846101 (C4 (Web Text))

Max Activating Token Index: 205

Text #11

Max Range: 3.3468. Min Range: -3.3468

Max Act: 3.1582. Min Act: -0.1700

Data Index: 59413 (C4 (Web Text))

Max Activating Token Index: 339

Text #12

Max Range: 3.3468. Min Range: -3.3468

Max Act: 3.1582. Min Act: -0.1700

Data Index: 1127748 (C4 (Web Text))

Max Activating Token Index: 102

Text #13

Max Range: 3.3468. Min Range: -3.3468

Max Act: 3.1582. Min Act: -0.1700

Data Index: 935745 (C4 (Web Text))

Max Activating Token Index: 425

Text #14

Max Range: 3.3468. Min Range: -3.3468

Max Act: 3.1582. Min Act: -0.1700

Data Index: 1204518 (C4 (Web Text))

Max Activating Token Index: 960

Text #15

Max Range: 3.3468. Min Range: -3.3468

Max Act: 3.1582. Min Act: -0.1700

Data Index: 1490594 (Python Code)

Max Activating Token Index: 643

Text #16

Max Range: 3.3468. Min Range: -3.3468

Max Act: 3.1109. Min Act: -0.1700

Data Index: 1355820 (C4 (Web Text))

Max Activating Token Index: 195

Text #17

Max Range: 3.3468. Min Range: -3.3468

Max Act: 3.1109. Min Act: -0.1700

Data Index: 558463 (C4 (Web Text))

Max Activating Token Index: 645

Text #18

Max Range: 3.3468. Min Range: -3.3468

Max Act: 3.1267. Min Act: -0.1700

Data Index: 1309060 (C4 (Web Text))

Max Activating Token Index: 1023

Text #19

Max Range: 3.3468. Min Range: -3.3468

Max Act: 3.1109. Min Act: -0.1700

Data Index: 453637 (C4 (Web Text))

Max Activating Token Index: 765