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

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

Neuron 18 in Layer 2

Load this data into an Interactive Neuroscope

See Documentation here

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

Text #0

Max Range: 5.7039. Min Range: -5.7039

Max Act: 5.7039. Min Act: -0.1700

Data Index: 1486023 (Python Code)

Max Activating Token Index: 517

Text #1

Max Range: 5.7039. Min Range: -5.7039

Max Act: 5.7039. Min Act: -0.1700

Data Index: 1620780 (Python Code)

Max Activating Token Index: 804

Text #2

Max Range: 5.7039. Min Range: -5.7039

Max Act: 5.5476. Min Act: -0.1700

Data Index: 1011977 (C4 (Web Text))

Max Activating Token Index: 799

Text #3

Max Range: 5.7039. Min Range: -5.7039

Max Act: 5.5789. Min Act: -0.1700

Data Index: 1504717 (Python Code)

Max Activating Token Index: 769

Text #4

Max Range: 5.7039. Min Range: -5.7039

Max Act: 5.6101. Min Act: -0.1700

Data Index: 1211125 (C4 (Web Text))

Max Activating Token Index: 54

Text #5

Max Range: 5.7039. Min Range: -5.7039

Max Act: 5.5789. Min Act: -0.1700

Data Index: 1402497 (Python Code)

Max Activating Token Index: 733

Text #6

Max Range: 5.7039. Min Range: -5.7039

Max Act: 5.3914. Min Act: -0.1700

Data Index: 1420628 (Python Code)

Max Activating Token Index: 469

Text #7

Max Range: 5.7039. Min Range: -5.7039

Max Act: 5.4539. Min Act: -0.1700

Data Index: 905441 (C4 (Web Text))

Max Activating Token Index: 622

Text #8

Max Range: 5.7039. Min Range: -5.7039

Max Act: 5.3914. Min Act: -0.1700

Data Index: 1564082 (Python Code)

Max Activating Token Index: 230

Text #9

Max Range: 5.7039. Min Range: -5.7039

Max Act: 5.4226. Min Act: -0.1700

Data Index: 1624205 (Python Code)

Max Activating Token Index: 1016

Text #10

Max Range: 5.7039. Min Range: -5.7039

Max Act: 5.3289. Min Act: -0.1700

Data Index: 1498197 (Python Code)

Max Activating Token Index: 203

Text #11

Max Range: 5.7039. Min Range: -5.7039

Max Act: 5.3601. Min Act: -0.1700

Data Index: 1365367 (Python Code)

Max Activating Token Index: 803

Text #12

Max Range: 5.7039. Min Range: -5.7039

Max Act: 5.3289. Min Act: -0.1700

Data Index: 1583038 (Python Code)

Max Activating Token Index: 127

Text #13

Max Range: 5.7039. Min Range: -5.7039

Max Act: 5.3601. Min Act: -0.1700

Data Index: 1519901 (Python Code)

Max Activating Token Index: 935

Text #14

Max Range: 5.7039. Min Range: -5.7039

Max Act: 5.2664. Min Act: -0.1700

Data Index: 1503207 (Python Code)

Max Activating Token Index: 513

Text #15

Max Range: 5.7039. Min Range: -5.7039

Max Act: 5.2976. Min Act: -0.1700

Data Index: 1412881 (Python Code)

Max Activating Token Index: 145

Text #16

Max Range: 5.7039. Min Range: -5.7039

Max Act: 5.3289. Min Act: -0.1700

Data Index: 441251 (C4 (Web Text))

Max Activating Token Index: 786

Text #17

Max Range: 5.7039. Min Range: -5.7039

Max Act: 5.2976. Min Act: -0.1700

Data Index: 1430015 (Python Code)

Max Activating Token Index: 890

Text #18

Max Range: 5.7039. Min Range: -5.7039

Max Act: 5.2664. Min Act: -0.1700

Data Index: 1582530 (Python Code)

Max Activating Token Index: 621

Text #19

Max Range: 5.7039. Min Range: -5.7039

Max Act: 5.3601. Min Act: -0.1700

Data Index: 1530196 (Python Code)

Max Activating Token Index: 489