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

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

Neuron 43 in Layer 0

Load this data into an Interactive Neuroscope

See Documentation here

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

Text #0

Max Range: 3.8915. Min Range: -3.8915

Max Act: 3.8915. Min Act: -0.1700

Data Index: 286442 (C4 (Web Text))

Max Activating Token Index: 7

Text #1

Max Range: 3.8915. Min Range: -3.8915

Max Act: 3.8915. Min Act: -0.1700

Data Index: 514763 (C4 (Web Text))

Max Activating Token Index: 6

Text #2

Max Range: 3.8915. Min Range: -3.8915

Max Act: 3.8602. Min Act: -0.1700

Data Index: 1638516 (Python Code)

Max Activating Token Index: 6

Text #3

Max Range: 3.8915. Min Range: -3.8915

Max Act: 3.8602. Min Act: -0.1700

Data Index: 449958 (C4 (Web Text))

Max Activating Token Index: 5

Text #4

Max Range: 3.8915. Min Range: -3.8915

Max Act: 3.8289. Min Act: -0.1700

Data Index: 666149 (C4 (Web Text))

Max Activating Token Index: 6

Text #5

Max Range: 3.8915. Min Range: -3.8915

Max Act: 3.8289. Min Act: -0.1700

Data Index: 135573 (C4 (Web Text))

Max Activating Token Index: 5

Text #6

Max Range: 3.8915. Min Range: -3.8915

Max Act: 3.8289. Min Act: -0.1700

Data Index: 383831 (C4 (Web Text))

Max Activating Token Index: 7

Text #7

Max Range: 3.8915. Min Range: -3.8915

Max Act: 3.8602. Min Act: -0.1700

Data Index: 32494 (C4 (Web Text))

Max Activating Token Index: 7

Text #8

Max Range: 3.8915. Min Range: -3.8915

Max Act: 3.8289. Min Act: -0.1700

Data Index: 169724 (C4 (Web Text))

Max Activating Token Index: 5

Text #9

Max Range: 3.8915. Min Range: -3.8915

Max Act: 3.8602. Min Act: -0.1700

Data Index: 1157398 (C4 (Web Text))

Max Activating Token Index: 6

Text #10

Max Range: 3.8915. Min Range: -3.8915

Max Act: 3.7977. Min Act: -0.1700

Data Index: 451192 (C4 (Web Text))

Max Activating Token Index: 7

Text #11

Max Range: 3.8915. Min Range: -3.8915

Max Act: 3.7977. Min Act: -0.1700

Data Index: 437547 (C4 (Web Text))

Max Activating Token Index: 7

Text #12

Max Range: 3.8915. Min Range: -3.8915

Max Act: 3.7977. Min Act: -0.1700

Data Index: 65559 (C4 (Web Text))

Max Activating Token Index: 6

Text #13

Max Range: 3.8915. Min Range: -3.8915

Max Act: 3.7977. Min Act: -0.1700

Data Index: 395538 (C4 (Web Text))

Max Activating Token Index: 5

Text #14

Max Range: 3.8915. Min Range: -3.8915

Max Act: 3.7977. Min Act: -0.1700

Data Index: 346279 (C4 (Web Text))

Max Activating Token Index: 6

Text #15

Max Range: 3.8915. Min Range: -3.8915

Max Act: 3.7977. Min Act: -0.1700

Data Index: 329333 (C4 (Web Text))

Max Activating Token Index: 5

Text #16

Max Range: 3.8915. Min Range: -3.8915

Max Act: 3.7977. Min Act: -0.1700

Data Index: 280085 (C4 (Web Text))

Max Activating Token Index: 6

Text #17

Max Range: 3.8915. Min Range: -3.8915

Max Act: 3.7977. Min Act: -0.1700

Data Index: 256505 (C4 (Web Text))

Max Activating Token Index: 7

Text #18

Max Range: 3.8915. Min Range: -3.8915

Max Act: 3.7977. Min Act: -0.1700

Data Index: 222549 (C4 (Web Text))

Max Activating Token Index: 6

Text #19

Max Range: 3.8915. Min Range: -3.8915

Max Act: 3.7977. Min Act: -0.1700

Data Index: 196939 (C4 (Web Text))

Max Activating Token Index: 3