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

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

Neuron 1582 in Layer 3

Load this data into an Interactive Neuroscope

See Documentation here

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

Text #0

Max Range: 8.0977. Min Range: -8.0977

Max Act: 8.0977. Min Act: -0.1700

Data Index: 1450172 (Python Code)

Max Activating Token Index: 259

Text #1

Max Range: 8.0977. Min Range: -8.0977

Max Act: 7.9727. Min Act: -0.1700

Data Index: 1498353 (Python Code)

Max Activating Token Index: 551

Text #2

Max Range: 8.0977. Min Range: -8.0977

Max Act: 7.8477. Min Act: -0.1700

Data Index: 1516179 (Python Code)

Max Activating Token Index: 197

Text #3

Max Range: 8.0977. Min Range: -8.0977

Max Act: 7.8789. Min Act: -0.1700

Data Index: 1429568 (Python Code)

Max Activating Token Index: 126

Text #4

Max Range: 8.0977. Min Range: -8.0977

Max Act: 7.8789. Min Act: -0.1700

Data Index: 1616894 (Python Code)

Max Activating Token Index: 659

Text #5

Max Range: 8.0977. Min Range: -8.0977

Max Act: 7.7852. Min Act: -0.1700

Data Index: 1391093 (Python Code)

Max Activating Token Index: 951

Text #6

Max Range: 8.0977. Min Range: -8.0977

Max Act: 7.7852. Min Act: -0.1700

Data Index: 1560922 (Python Code)

Max Activating Token Index: 815

Text #7

Max Range: 8.0977. Min Range: -8.0977

Max Act: 7.7852. Min Act: -0.1700

Data Index: 1592600 (Python Code)

Max Activating Token Index: 407

Text #8

Max Range: 8.0977. Min Range: -8.0977

Max Act: 7.7539. Min Act: -0.1700

Data Index: 1525395 (Python Code)

Max Activating Token Index: 333

Text #9

Max Range: 8.0977. Min Range: -8.0977

Max Act: 7.6602. Min Act: -0.1700

Data Index: 1394542 (Python Code)

Max Activating Token Index: 961

Text #10

Max Range: 8.0977. Min Range: -8.0977

Max Act: 7.6914. Min Act: -0.1700

Data Index: 1577916 (Python Code)

Max Activating Token Index: 956

Text #11

Max Range: 8.0977. Min Range: -8.0977

Max Act: 7.6914. Min Act: -0.1700

Data Index: 1397791 (Python Code)

Max Activating Token Index: 169

Text #12

Max Range: 8.0977. Min Range: -8.0977

Max Act: 7.6289. Min Act: -0.1700

Data Index: 1642790 (Python Code)

Max Activating Token Index: 372

Text #13

Max Range: 8.0977. Min Range: -8.0977

Max Act: 7.5664. Min Act: -0.1700

Data Index: 1386849 (Python Code)

Max Activating Token Index: 550

Text #14

Max Range: 8.0977. Min Range: -8.0977

Max Act: 7.5664. Min Act: -0.1700

Data Index: 1639674 (Python Code)

Max Activating Token Index: 363

Text #15

Max Range: 8.0977. Min Range: -8.0977

Max Act: 7.5977. Min Act: -0.1700

Data Index: 1511208 (Python Code)

Max Activating Token Index: 408

Text #16

Max Range: 8.0977. Min Range: -8.0977

Max Act: 7.5039. Min Act: -0.1700

Data Index: 1429769 (Python Code)

Max Activating Token Index: 984

Text #17

Max Range: 8.0977. Min Range: -8.0977

Max Act: 7.5352. Min Act: -0.1700

Data Index: 1626480 (Python Code)

Max Activating Token Index: 162

Text #18

Max Range: 8.0977. Min Range: -8.0977

Max Act: 7.5039. Min Act: -0.1700

Data Index: 1454945 (Python Code)

Max Activating Token Index: 182

Text #19

Max Range: 8.0977. Min Range: -8.0977

Max Act: 7.5352. Min Act: -0.1700

Data Index: 1583688 (Python Code)

Max Activating Token Index: 174