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

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

Neuron 1500 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: 7.3186. Min Range: -7.3186

Max Act: 7.3186. Min Act: -0.1700

Data Index: 1327113 (C4 (Web Text))

Max Activating Token Index: 322

Text #1

Max Range: 7.3186. Min Range: -7.3186

Max Act: 7.1623. Min Act: -0.1700

Data Index: 85913 (C4 (Web Text))

Max Activating Token Index: 69

Text #2

Max Range: 7.3186. Min Range: -7.3186

Max Act: 7.0061. Min Act: -0.1700

Data Index: 507997 (C4 (Web Text))

Max Activating Token Index: 348

Text #3

Max Range: 7.3186. Min Range: -7.3186

Max Act: 6.8498. Min Act: -0.1700

Data Index: 1175269 (C4 (Web Text))

Max Activating Token Index: 502

Text #4

Max Range: 7.3186. Min Range: -7.3186

Max Act: 6.7873. Min Act: -0.1700

Data Index: 3923 (C4 (Web Text))

Max Activating Token Index: 490

Text #5

Max Range: 7.3186. Min Range: -7.3186

Max Act: 6.7248. Min Act: -0.1700

Data Index: 9051 (C4 (Web Text))

Max Activating Token Index: 412

Text #6

Max Range: 7.3186. Min Range: -7.3186

Max Act: 6.5686. Min Act: -0.1700

Data Index: 44378 (C4 (Web Text))

Max Activating Token Index: 487

Text #7

Max Range: 7.3186. Min Range: -7.3186

Max Act: 6.6311. Min Act: -0.1700

Data Index: 1248567 (C4 (Web Text))

Max Activating Token Index: 119

Text #8

Max Range: 7.3186. Min Range: -7.3186

Max Act: 6.5998. Min Act: -0.1700

Data Index: 289193 (C4 (Web Text))

Max Activating Token Index: 559

Text #9

Max Range: 7.3186. Min Range: -7.3186

Max Act: 6.6936. Min Act: -0.1700

Data Index: 595064 (C4 (Web Text))

Max Activating Token Index: 729

Text #10

Max Range: 7.3186. Min Range: -7.3186

Max Act: 6.5686. Min Act: -0.1700

Data Index: 724568 (C4 (Web Text))

Max Activating Token Index: 628

Text #11

Max Range: 7.3186. Min Range: -7.3186

Max Act: 6.5998. Min Act: -0.1700

Data Index: 207555 (C4 (Web Text))

Max Activating Token Index: 138

Text #12

Max Range: 7.3186. Min Range: -7.3186

Max Act: 6.5998. Min Act: -0.1700

Data Index: 430716 (C4 (Web Text))

Max Activating Token Index: 592

Text #13

Max Range: 7.3186. Min Range: -7.3186

Max Act: 6.5686. Min Act: -0.1700

Data Index: 679603 (C4 (Web Text))

Max Activating Token Index: 1001

Text #14

Max Range: 7.3186. Min Range: -7.3186

Max Act: 6.5998. Min Act: -0.1700

Data Index: 79203 (C4 (Web Text))

Max Activating Token Index: 640

Text #15

Max Range: 7.3186. Min Range: -7.3186

Max Act: 6.6311. Min Act: -0.1700

Data Index: 888846 (C4 (Web Text))

Max Activating Token Index: 903

Text #16

Max Range: 7.3186. Min Range: -7.3186

Max Act: 6.4748. Min Act: -0.1700

Data Index: 399694 (C4 (Web Text))

Max Activating Token Index: 775

Text #17

Max Range: 7.3186. Min Range: -7.3186

Max Act: 6.5686. Min Act: -0.1700

Data Index: 842823 (C4 (Web Text))

Max Activating Token Index: 363

Text #18

Max Range: 7.3186. Min Range: -7.3186

Max Act: 6.5686. Min Act: -0.1700

Data Index: 61377 (C4 (Web Text))

Max Activating Token Index: 234

Text #19

Max Range: 7.3186. Min Range: -7.3186

Max Act: 6.4748. Min Act: -0.1700

Data Index: 1049215 (C4 (Web Text))

Max Activating Token Index: 292