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

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

Neuron 1000 in Layer 0

Load this data into an Interactive Neuroscope

See Documentation here

Transformer Lens Loading: HookedTransformer.from_pretrained('gelu-3l')

Text #0

Max Range: 3.4284. Min Range: -3.4284

Max Act: 3.4284. Min Act: -0.1700

Data Index: 1192392 (C4 (Web Text))

Max Activating Token Index: 76

Text #1

Max Range: 3.4284. Min Range: -3.4284

Max Act: 3.4127. Min Act: -0.1700

Data Index: 742998 (C4 (Web Text))

Max Activating Token Index: 390

Text #2

Max Range: 3.4284. Min Range: -3.4284

Max Act: 3.3813. Min Act: -0.1700

Data Index: 1258999 (C4 (Web Text))

Max Activating Token Index: 90

Text #3

Max Range: 3.4284. Min Range: -3.4284

Max Act: 3.3813. Min Act: -0.1700

Data Index: 768167 (C4 (Web Text))

Max Activating Token Index: 153

Text #4

Max Range: 3.4284. Min Range: -3.4284

Max Act: 3.3970. Min Act: -0.1700

Data Index: 1280286 (C4 (Web Text))

Max Activating Token Index: 878

Text #5

Max Range: 3.4284. Min Range: -3.4284

Max Act: 3.3656. Min Act: -0.1700

Data Index: 646234 (C4 (Web Text))

Max Activating Token Index: 252

Text #6

Max Range: 3.4284. Min Range: -3.4284

Max Act: 3.3656. Min Act: -0.1700

Data Index: 394898 (C4 (Web Text))

Max Activating Token Index: 424

Text #7

Max Range: 3.4284. Min Range: -3.4284

Max Act: 3.3656. Min Act: -0.1700

Data Index: 1016162 (C4 (Web Text))

Max Activating Token Index: 145

Text #8

Max Range: 3.4284. Min Range: -3.4284

Max Act: 3.3499. Min Act: -0.1700

Data Index: 493407 (C4 (Web Text))

Max Activating Token Index: 978

Text #9

Max Range: 3.4284. Min Range: -3.4284

Max Act: 3.3342. Min Act: -0.1700

Data Index: 665852 (C4 (Web Text))

Max Activating Token Index: 106

Text #10

Max Range: 3.4284. Min Range: -3.4284

Max Act: 3.3342. Min Act: -0.1700

Data Index: 420996 (C4 (Web Text))

Max Activating Token Index: 165

Text #11

Max Range: 3.4284. Min Range: -3.4284

Max Act: 3.3499. Min Act: -0.1700

Data Index: 116266 (C4 (Web Text))

Max Activating Token Index: 633

Text #12

Max Range: 3.4284. Min Range: -3.4284

Max Act: 3.3342. Min Act: -0.1700

Data Index: 689973 (C4 (Web Text))

Max Activating Token Index: 210

Text #13

Max Range: 3.4284. Min Range: -3.4284

Max Act: 3.3342. Min Act: -0.1700

Data Index: 600315 (C4 (Web Text))

Max Activating Token Index: 61

Text #14

Max Range: 3.4284. Min Range: -3.4284

Max Act: 3.3185. Min Act: -0.1700

Data Index: 1324314 (C4 (Web Text))

Max Activating Token Index: 306

Text #15

Max Range: 3.4284. Min Range: -3.4284

Max Act: 3.3028. Min Act: -0.1700

Data Index: 346166 (C4 (Web Text))

Max Activating Token Index: 234

Text #16

Max Range: 3.4284. Min Range: -3.4284

Max Act: 3.3185. Min Act: -0.1700

Data Index: 95627 (C4 (Web Text))

Max Activating Token Index: 254

Text #17

Max Range: 3.4284. Min Range: -3.4284

Max Act: 3.3028. Min Act: -0.1700

Data Index: 863700 (C4 (Web Text))

Max Activating Token Index: 7

Text #18

Max Range: 3.4284. Min Range: -3.4284

Max Act: 3.3028. Min Act: -0.1700

Data Index: 319713 (C4 (Web Text))

Max Activating Token Index: 318

Text #19

Max Range: 3.4284. Min Range: -3.4284

Max Act: 3.2871. Min Act: -0.1700

Data Index: 488094 (C4 (Web Text))

Max Activating Token Index: 86