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

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

Neuron 1117 in Layer 1

Load this data into an Interactive Neuroscope

See Documentation here

Transformer Lens Loading: HookedTransformer.from_pretrained('gelu-4l')

Text #0

Max Range: 3.2965. Min Range: -3.2965

Max Act: 3.2965. Min Act: -0.1700

Data Index: 462049 (C4 (Web Text))

Max Activating Token Index: 865

Text #1

Max Range: 3.2965. Min Range: -3.2965

Max Act: 3.2651. Min Act: -0.1700

Data Index: 978667 (C4 (Web Text))

Max Activating Token Index: 571

Text #2

Max Range: 3.2965. Min Range: -3.2965

Max Act: 3.2179. Min Act: -0.1700

Data Index: 1220004 (C4 (Web Text))

Max Activating Token Index: 612

Text #3

Max Range: 3.2965. Min Range: -3.2965

Max Act: 3.2022. Min Act: -0.1700

Data Index: 272784 (C4 (Web Text))

Max Activating Token Index: 129

Text #4

Max Range: 3.2965. Min Range: -3.2965

Max Act: 3.2022. Min Act: -0.1700

Data Index: 1307975 (C4 (Web Text))

Max Activating Token Index: 278

Text #5

Max Range: 3.2965. Min Range: -3.2965

Max Act: 3.1864. Min Act: -0.1700

Data Index: 442795 (C4 (Web Text))

Max Activating Token Index: 574

Text #6

Max Range: 3.2965. Min Range: -3.2965

Max Act: 3.1707. Min Act: -0.1700

Data Index: 1287572 (C4 (Web Text))

Max Activating Token Index: 878

Text #7

Max Range: 3.2965. Min Range: -3.2965

Max Act: 3.1392. Min Act: -0.1700

Data Index: 1368728 (Python Code)

Max Activating Token Index: 513

Text #8

Max Range: 3.2965. Min Range: -3.2965

Max Act: 3.1392. Min Act: -0.1700

Data Index: 715788 (C4 (Web Text))

Max Activating Token Index: 1010

Text #9

Max Range: 3.2965. Min Range: -3.2965

Max Act: 3.1077. Min Act: -0.1700

Data Index: 469616 (C4 (Web Text))

Max Activating Token Index: 276

Text #10

Max Range: 3.2965. Min Range: -3.2965

Max Act: 3.1077. Min Act: -0.1700

Data Index: 1635131 (Python Code)

Max Activating Token Index: 160

Text #11

Max Range: 3.2965. Min Range: -3.2965

Max Act: 3.0604. Min Act: -0.1700

Data Index: 1592590 (Python Code)

Max Activating Token Index: 504

Text #12

Max Range: 3.2965. Min Range: -3.2965

Max Act: 3.0446. Min Act: -0.1700

Data Index: 151432 (C4 (Web Text))

Max Activating Token Index: 16

Text #13

Max Range: 3.2965. Min Range: -3.2965

Max Act: 3.0288. Min Act: -0.1700

Data Index: 983660 (C4 (Web Text))

Max Activating Token Index: 1023

Text #14

Max Range: 3.2965. Min Range: -3.2965

Max Act: 3.0446. Min Act: -0.1700

Data Index: 942125 (C4 (Web Text))

Max Activating Token Index: 21

Text #15

Max Range: 3.2965. Min Range: -3.2965

Max Act: 3.0604. Min Act: -0.1700

Data Index: 659557 (C4 (Web Text))

Max Activating Token Index: 78

Text #16

Max Range: 3.2965. Min Range: -3.2965

Max Act: 3.0288. Min Act: -0.1700

Data Index: 180496 (C4 (Web Text))

Max Activating Token Index: 76

Text #17

Max Range: 3.2965. Min Range: -3.2965

Max Act: 3.0130. Min Act: -0.1700

Data Index: 863983 (C4 (Web Text))

Max Activating Token Index: 162

Text #18

Max Range: 3.2965. Min Range: -3.2965

Max Act: 3.0130. Min Act: -0.1700

Data Index: 1497389 (Python Code)

Max Activating Token Index: 736

Text #19

Max Range: 3.2965. Min Range: -3.2965

Max Act: 3.0130. Min Act: -0.1700

Data Index: 1628777 (Python Code)

Max Activating Token Index: 155