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

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

Neuron 2004 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: 5.2909. Min Range: -5.2909

Max Act: 5.2909. Min Act: -0.1700

Data Index: 1139326 (C4 (Web Text))

Max Activating Token Index: 180

Text #1

Max Range: 5.2909. Min Range: -5.2909

Max Act: 5.0096. Min Act: -0.1700

Data Index: 486485 (C4 (Web Text))

Max Activating Token Index: 990

Text #2

Max Range: 5.2909. Min Range: -5.2909

Max Act: 5.0409. Min Act: -0.1700

Data Index: 1282513 (C4 (Web Text))

Max Activating Token Index: 1011

Text #3

Max Range: 5.2909. Min Range: -5.2909

Max Act: 5.0409. Min Act: -0.1700

Data Index: 993547 (C4 (Web Text))

Max Activating Token Index: 314

Text #4

Max Range: 5.2909. Min Range: -5.2909

Max Act: 4.9783. Min Act: -0.1696

Data Index: 76116 (C4 (Web Text))

Max Activating Token Index: 49

Text #5

Max Range: 5.2909. Min Range: -5.2909

Max Act: 4.9471. Min Act: -0.1700

Data Index: 628037 (C4 (Web Text))

Max Activating Token Index: 737

Text #6

Max Range: 5.2909. Min Range: -5.2909

Max Act: 4.8846. Min Act: -0.1700

Data Index: 1159329 (C4 (Web Text))

Max Activating Token Index: 1008

Text #7

Max Range: 5.2909. Min Range: -5.2909

Max Act: 4.8846. Min Act: -0.1699

Data Index: 1250451 (C4 (Web Text))

Max Activating Token Index: 145

Text #8

Max Range: 5.2909. Min Range: -5.2909

Max Act: 4.8846. Min Act: -0.1700

Data Index: 296040 (C4 (Web Text))

Max Activating Token Index: 521

Text #9

Max Range: 5.2909. Min Range: -5.2909

Max Act: 4.8221. Min Act: -0.1700

Data Index: 390839 (C4 (Web Text))

Max Activating Token Index: 788

Text #10

Max Range: 5.2909. Min Range: -5.2909

Max Act: 4.7908. Min Act: -0.1699

Data Index: 499391 (C4 (Web Text))

Max Activating Token Index: 61

Text #11

Max Range: 5.2909. Min Range: -5.2909

Max Act: 4.8533. Min Act: -0.1700

Data Index: 29559 (C4 (Web Text))

Max Activating Token Index: 190

Text #12

Max Range: 5.2909. Min Range: -5.2909

Max Act: 4.7908. Min Act: -0.1700

Data Index: 1060779 (C4 (Web Text))

Max Activating Token Index: 375

Text #13

Max Range: 5.2909. Min Range: -5.2909

Max Act: 4.7908. Min Act: -0.1700

Data Index: 938720 (C4 (Web Text))

Max Activating Token Index: 1019

Text #14

Max Range: 5.2909. Min Range: -5.2909

Max Act: 4.8221. Min Act: -0.1700

Data Index: 601381 (C4 (Web Text))

Max Activating Token Index: 904

Text #15

Max Range: 5.2909. Min Range: -5.2909

Max Act: 4.7908. Min Act: -0.1700

Data Index: 704498 (C4 (Web Text))

Max Activating Token Index: 294

Text #16

Max Range: 5.2909. Min Range: -5.2909

Max Act: 4.8221. Min Act: -0.1700

Data Index: 262249 (C4 (Web Text))

Max Activating Token Index: 170

Text #17

Max Range: 5.2909. Min Range: -5.2909

Max Act: 4.7596. Min Act: -0.1700

Data Index: 271049 (C4 (Web Text))

Max Activating Token Index: 835

Text #18

Max Range: 5.2909. Min Range: -5.2909

Max Act: 4.7596. Min Act: -0.1700

Data Index: 644359 (C4 (Web Text))

Max Activating Token Index: 649

Text #19

Max Range: 5.2909. Min Range: -5.2909

Max Act: 4.7596. Min Act: -0.1700

Data Index: 1121230 (C4 (Web Text))

Max Activating Token Index: 339