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

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

Neuron 635 in Layer 0

Load this data into an Interactive Neuroscope

See Documentation here

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

Text #0

Max Range: 3.4947. Min Range: -3.4947

Max Act: 3.4947. Min Act: -0.1700

Data Index: 1587635 (Python Code)

Max Activating Token Index: 9

Text #1

Max Range: 3.4947. Min Range: -3.4947

Max Act: 3.4790. Min Act: -0.1700

Data Index: 1545035 (Python Code)

Max Activating Token Index: 624

Text #2

Max Range: 3.4947. Min Range: -3.4947

Max Act: 3.4633. Min Act: -0.1700

Data Index: 1468775 (Python Code)

Max Activating Token Index: 572

Text #3

Max Range: 3.4947. Min Range: -3.4947

Max Act: 3.4320. Min Act: -0.1700

Data Index: 556955 (C4 (Web Text))

Max Activating Token Index: 370

Text #4

Max Range: 3.4947. Min Range: -3.4947

Max Act: 3.4163. Min Act: -0.1700

Data Index: 1526517 (Python Code)

Max Activating Token Index: 180

Text #5

Max Range: 3.4947. Min Range: -3.4947

Max Act: 3.4163. Min Act: -0.1700

Data Index: 1373882 (Python Code)

Max Activating Token Index: 28

Text #6

Max Range: 3.4947. Min Range: -3.4947

Max Act: 3.4320. Min Act: -0.1700

Data Index: 1652347 (Python Code)

Max Activating Token Index: 9

Text #7

Max Range: 3.4947. Min Range: -3.4947

Max Act: 3.4163. Min Act: -0.1700

Data Index: 1545506 (Python Code)

Max Activating Token Index: 758

Text #8

Max Range: 3.4947. Min Range: -3.4947

Max Act: 3.4163. Min Act: -0.1700

Data Index: 1560978 (Python Code)

Max Activating Token Index: 546

Text #9

Max Range: 3.4947. Min Range: -3.4947

Max Act: 3.3849. Min Act: -0.1700

Data Index: 1620083 (Python Code)

Max Activating Token Index: 4

Text #10

Max Range: 3.4947. Min Range: -3.4947

Max Act: 3.3849. Min Act: -0.1700

Data Index: 1626172 (Python Code)

Max Activating Token Index: 4

Text #11

Max Range: 3.4947. Min Range: -3.4947

Max Act: 3.3849. Min Act: -0.1700

Data Index: 1652238 (Python Code)

Max Activating Token Index: 4

Text #12

Max Range: 3.4947. Min Range: -3.4947

Max Act: 3.3849. Min Act: -0.1700

Data Index: 1554076 (Python Code)

Max Activating Token Index: 4

Text #13

Max Range: 3.4947. Min Range: -3.4947

Max Act: 3.3849. Min Act: -0.1700

Data Index: 1395410 (Python Code)

Max Activating Token Index: 4

Text #14

Max Range: 3.4947. Min Range: -3.4947

Max Act: 3.4006. Min Act: -0.1700

Data Index: 1554990 (Python Code)

Max Activating Token Index: 17

Text #15

Max Range: 3.4947. Min Range: -3.4947

Max Act: 3.3849. Min Act: -0.1700

Data Index: 1478262 (Python Code)

Max Activating Token Index: 115

Text #16

Max Range: 3.4947. Min Range: -3.4947

Max Act: 3.3849. Min Act: -0.1700

Data Index: 1583051 (Python Code)

Max Activating Token Index: 252

Text #17

Max Range: 3.4947. Min Range: -3.4947

Max Act: 3.3535. Min Act: -0.1700

Data Index: 1628765 (Python Code)

Max Activating Token Index: 48

Text #18

Max Range: 3.4947. Min Range: -3.4947

Max Act: 3.3692. Min Act: -0.1700

Data Index: 1603426 (Python Code)

Max Activating Token Index: 7

Text #19

Max Range: 3.4947. Min Range: -3.4947

Max Act: 3.3535. Min Act: -0.1700

Data Index: 1572229 (Python Code)

Max Activating Token Index: 4