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

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

Neuron 389 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: 3.5593. Min Range: -3.5593

Max Act: 3.5593. Min Act: -0.1700

Data Index: 772401 (C4 (Web Text))

Max Activating Token Index: 711

Text #1

Max Range: 3.5593. Min Range: -3.5593

Max Act: 3.5593. Min Act: -0.1700

Data Index: 452972 (C4 (Web Text))

Max Activating Token Index: 588

Text #2

Max Range: 3.5593. Min Range: -3.5593

Max Act: 3.4967. Min Act: -0.1700

Data Index: 1044522 (C4 (Web Text))

Max Activating Token Index: 721

Text #3

Max Range: 3.5593. Min Range: -3.5593

Max Act: 3.5280. Min Act: -0.1700

Data Index: 1617814 (Python Code)

Max Activating Token Index: 988

Text #4

Max Range: 3.5593. Min Range: -3.5593

Max Act: 3.5280. Min Act: -0.1700

Data Index: 1314950 (C4 (Web Text))

Max Activating Token Index: 51

Text #5

Max Range: 3.5593. Min Range: -3.5593

Max Act: 3.4967. Min Act: -0.1700

Data Index: 1499137 (Python Code)

Max Activating Token Index: 1009

Text #6

Max Range: 3.5593. Min Range: -3.5593

Max Act: 3.4653. Min Act: -0.1700

Data Index: 439187 (C4 (Web Text))

Max Activating Token Index: 722

Text #7

Max Range: 3.5593. Min Range: -3.5593

Max Act: 3.4653. Min Act: -0.1700

Data Index: 730643 (C4 (Web Text))

Max Activating Token Index: 945

Text #8

Max Range: 3.5593. Min Range: -3.5593

Max Act: 3.4653. Min Act: -0.1700

Data Index: 1606935 (Python Code)

Max Activating Token Index: 399

Text #9

Max Range: 3.5593. Min Range: -3.5593

Max Act: 3.4653. Min Act: -0.1700

Data Index: 1375643 (Python Code)

Max Activating Token Index: 969

Text #10

Max Range: 3.5593. Min Range: -3.5593

Max Act: 3.4340. Min Act: -0.1700

Data Index: 1376016 (Python Code)

Max Activating Token Index: 967

Text #11

Max Range: 3.5593. Min Range: -3.5593

Max Act: 3.4026. Min Act: -0.1700

Data Index: 1583851 (Python Code)

Max Activating Token Index: 591

Text #12

Max Range: 3.5593. Min Range: -3.5593

Max Act: 3.4340. Min Act: -0.1700

Data Index: 1059701 (C4 (Web Text))

Max Activating Token Index: 943

Text #13

Max Range: 3.5593. Min Range: -3.5593

Max Act: 3.4340. Min Act: -0.1700

Data Index: 266541 (C4 (Web Text))

Max Activating Token Index: 16

Text #14

Max Range: 3.5593. Min Range: -3.5593

Max Act: 3.4340. Min Act: -0.1700

Data Index: 462588 (C4 (Web Text))

Max Activating Token Index: 838

Text #15

Max Range: 3.5593. Min Range: -3.5593

Max Act: 3.4026. Min Act: -0.1700

Data Index: 987379 (C4 (Web Text))

Max Activating Token Index: 419

Text #16

Max Range: 3.5593. Min Range: -3.5593

Max Act: 3.4026. Min Act: -0.1700

Data Index: 989739 (C4 (Web Text))

Max Activating Token Index: 578

Text #17

Max Range: 3.5593. Min Range: -3.5593

Max Act: 3.3869. Min Act: -0.1700

Data Index: 652160 (C4 (Web Text))

Max Activating Token Index: 675

Text #18

Max Range: 3.5593. Min Range: -3.5593

Max Act: 3.4026. Min Act: -0.1700

Data Index: 300757 (C4 (Web Text))

Max Activating Token Index: 828

Text #19

Max Range: 3.5593. Min Range: -3.5593

Max Act: 3.4026. Min Act: -0.1700

Data Index: 1349469 (C4 (Web Text))

Max Activating Token Index: 785