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

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

Neuron 25 in Layer 1

Load this data into an Interactive Neuroscope

See Documentation here

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

Text #0

Max Range: 3.3914. Min Range: -3.3914

Max Act: 3.3914. Min Act: -0.1700

Data Index: 1016856 (C4 (Web Text))

Max Activating Token Index: 191

Text #1

Max Range: 3.3914. Min Range: -3.3914

Max Act: 3.3286. Min Act: -0.1700

Data Index: 201237 (C4 (Web Text))

Max Activating Token Index: 968

Text #2

Max Range: 3.3914. Min Range: -3.3914

Max Act: 3.2972. Min Act: -0.1700

Data Index: 1038688 (C4 (Web Text))

Max Activating Token Index: 454

Text #3

Max Range: 3.3914. Min Range: -3.3914

Max Act: 3.2972. Min Act: -0.1700

Data Index: 312774 (C4 (Web Text))

Max Activating Token Index: 452

Text #4

Max Range: 3.3914. Min Range: -3.3914

Max Act: 3.3286. Min Act: -0.1700

Data Index: 789921 (C4 (Web Text))

Max Activating Token Index: 628

Text #5

Max Range: 3.3914. Min Range: -3.3914

Max Act: 3.2972. Min Act: -0.1700

Data Index: 294641 (C4 (Web Text))

Max Activating Token Index: 833

Text #6

Max Range: 3.3914. Min Range: -3.3914

Max Act: 3.2972. Min Act: -0.1700

Data Index: 324910 (C4 (Web Text))

Max Activating Token Index: 338

Text #7

Max Range: 3.3914. Min Range: -3.3914

Max Act: 3.2972. Min Act: -0.1700

Data Index: 670981 (C4 (Web Text))

Max Activating Token Index: 278

Text #8

Max Range: 3.3914. Min Range: -3.3914

Max Act: 3.2815. Min Act: -0.1700

Data Index: 701859 (C4 (Web Text))

Max Activating Token Index: 756

Text #9

Max Range: 3.3914. Min Range: -3.3914

Max Act: 3.2658. Min Act: -0.1700

Data Index: 812943 (C4 (Web Text))

Max Activating Token Index: 631

Text #10

Max Range: 3.3914. Min Range: -3.3914

Max Act: 3.2501. Min Act: -0.1700

Data Index: 1090246 (C4 (Web Text))

Max Activating Token Index: 887

Text #11

Max Range: 3.3914. Min Range: -3.3914

Max Act: 3.2344. Min Act: -0.1700

Data Index: 1346782 (C4 (Web Text))

Max Activating Token Index: 984

Text #12

Max Range: 3.3914. Min Range: -3.3914

Max Act: 3.2029. Min Act: -0.1700

Data Index: 176039 (C4 (Web Text))

Max Activating Token Index: 1006

Text #13

Max Range: 3.3914. Min Range: -3.3914

Max Act: 3.2344. Min Act: -0.1700

Data Index: 1011802 (C4 (Web Text))

Max Activating Token Index: 871

Text #14

Max Range: 3.3914. Min Range: -3.3914

Max Act: 3.2501. Min Act: -0.1700

Data Index: 441816 (C4 (Web Text))

Max Activating Token Index: 459

Text #15

Max Range: 3.3914. Min Range: -3.3914

Max Act: 3.2344. Min Act: -0.1700

Data Index: 908585 (C4 (Web Text))

Max Activating Token Index: 323

Text #16

Max Range: 3.3914. Min Range: -3.3914

Max Act: 3.2344. Min Act: -0.1700

Data Index: 12148 (C4 (Web Text))

Max Activating Token Index: 426

Text #17

Max Range: 3.3914. Min Range: -3.3914

Max Act: 3.2186. Min Act: -0.1700

Data Index: 1209443 (C4 (Web Text))

Max Activating Token Index: 939

Text #18

Max Range: 3.3914. Min Range: -3.3914

Max Act: 3.2029. Min Act: -0.1700

Data Index: 995066 (C4 (Web Text))

Max Activating Token Index: 702

Text #19

Max Range: 3.3914. Min Range: -3.3914

Max Act: 3.2186. Min Act: -0.1700

Data Index: 60254 (C4 (Web Text))

Max Activating Token Index: 251