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

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

Neuron 1047 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: 4.1401. Min Range: -4.1401

Max Act: 4.1401. Min Act: -0.1700

Data Index: 529658 (C4 (Web Text))

Max Activating Token Index: 958

Text #1

Max Range: 4.1401. Min Range: -4.1401

Max Act: 4.1244. Min Act: -0.1700

Data Index: 196057 (C4 (Web Text))

Max Activating Token Index: 348

Text #2

Max Range: 4.1401. Min Range: -4.1401

Max Act: 4.0776. Min Act: -0.1700

Data Index: 353904 (C4 (Web Text))

Max Activating Token Index: 391

Text #3

Max Range: 4.1401. Min Range: -4.1401

Max Act: 4.0776. Min Act: -0.1700

Data Index: 735134 (C4 (Web Text))

Max Activating Token Index: 494

Text #4

Max Range: 4.1401. Min Range: -4.1401

Max Act: 3.9994. Min Act: -0.1700

Data Index: 1200419 (C4 (Web Text))

Max Activating Token Index: 309

Text #5

Max Range: 4.1401. Min Range: -4.1401

Max Act: 3.9838. Min Act: -0.1700

Data Index: 672405 (C4 (Web Text))

Max Activating Token Index: 400

Text #6

Max Range: 4.1401. Min Range: -4.1401

Max Act: 3.9525. Min Act: -0.1700

Data Index: 1263978 (C4 (Web Text))

Max Activating Token Index: 891

Text #7

Max Range: 4.1401. Min Range: -4.1401

Max Act: 3.9056. Min Act: -0.1700

Data Index: 542933 (C4 (Web Text))

Max Activating Token Index: 944

Text #8

Max Range: 4.1401. Min Range: -4.1401

Max Act: 3.9212. Min Act: -0.1700

Data Index: 1333010 (C4 (Web Text))

Max Activating Token Index: 947

Text #9

Max Range: 4.1401. Min Range: -4.1401

Max Act: 3.8900. Min Act: -0.1700

Data Index: 691234 (C4 (Web Text))

Max Activating Token Index: 189

Text #10

Max Range: 4.1401. Min Range: -4.1401

Max Act: 3.8430. Min Act: -0.1700

Data Index: 669222 (C4 (Web Text))

Max Activating Token Index: 682

Text #11

Max Range: 4.1401. Min Range: -4.1401

Max Act: 3.8900. Min Act: -0.1700

Data Index: 1116319 (C4 (Web Text))

Max Activating Token Index: 971

Text #12

Max Range: 4.1401. Min Range: -4.1401

Max Act: 3.8743. Min Act: -0.1700

Data Index: 660561 (C4 (Web Text))

Max Activating Token Index: 754

Text #13

Max Range: 4.1401. Min Range: -4.1401

Max Act: 3.9056. Min Act: -0.1700

Data Index: 597294 (C4 (Web Text))

Max Activating Token Index: 1003

Text #14

Max Range: 4.1401. Min Range: -4.1401

Max Act: 3.8430. Min Act: -0.1700

Data Index: 102339 (C4 (Web Text))

Max Activating Token Index: 546

Text #15

Max Range: 4.1401. Min Range: -4.1401

Max Act: 3.8118. Min Act: -0.1700

Data Index: 220338 (C4 (Web Text))

Max Activating Token Index: 128

Text #16

Max Range: 4.1401. Min Range: -4.1401

Max Act: 3.8430. Min Act: -0.1700

Data Index: 226924 (C4 (Web Text))

Max Activating Token Index: 576

Text #17

Max Range: 4.1401. Min Range: -4.1401

Max Act: 3.8587. Min Act: -0.1700

Data Index: 988030 (C4 (Web Text))

Max Activating Token Index: 983

Text #18

Max Range: 4.1401. Min Range: -4.1401

Max Act: 3.8430. Min Act: -0.1700

Data Index: 170324 (C4 (Web Text))

Max Activating Token Index: 118

Text #19

Max Range: 4.1401. Min Range: -4.1401

Max Act: 3.8274. Min Act: -0.1700

Data Index: 480614 (C4 (Web Text))

Max Activating Token Index: 30