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

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

Neuron 779 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: 5.5950. Min Range: -5.5950

Max Act: 5.5950. Min Act: -0.1699

Data Index: 468802 (C4 (Web Text))

Max Activating Token Index: 477

Text #1

Max Range: 5.5950. Min Range: -5.5950

Max Act: 5.4700. Min Act: -0.1371

Data Index: 1627354 (Python Code)

Max Activating Token Index: 844

Text #2

Max Range: 5.5950. Min Range: -5.5950

Max Act: 5.3450. Min Act: -0.1700

Data Index: 87463 (C4 (Web Text))

Max Activating Token Index: 502

Text #3

Max Range: 5.5950. Min Range: -5.5950

Max Act: 5.2513. Min Act: -0.1700

Data Index: 1616653 (Python Code)

Max Activating Token Index: 217

Text #4

Max Range: 5.5950. Min Range: -5.5950

Max Act: 5.3450. Min Act: -0.1700

Data Index: 1515639 (Python Code)

Max Activating Token Index: 694

Text #5

Max Range: 5.5950. Min Range: -5.5950

Max Act: 5.1888. Min Act: -0.1700

Data Index: 1439840 (Python Code)

Max Activating Token Index: 757

Text #6

Max Range: 5.5950. Min Range: -5.5950

Max Act: 5.2200. Min Act: -0.1700

Data Index: 273762 (C4 (Web Text))

Max Activating Token Index: 329

Text #7

Max Range: 5.5950. Min Range: -5.5950

Max Act: 5.1888. Min Act: -0.1700

Data Index: 1001974 (C4 (Web Text))

Max Activating Token Index: 696

Text #8

Max Range: 5.5950. Min Range: -5.5950

Max Act: 5.0950. Min Act: -0.1700

Data Index: 328466 (C4 (Web Text))

Max Activating Token Index: 444

Text #9

Max Range: 5.5950. Min Range: -5.5950

Max Act: 5.0950. Min Act: -0.1700

Data Index: 1504927 (Python Code)

Max Activating Token Index: 846

Text #10

Max Range: 5.5950. Min Range: -5.5950

Max Act: 5.0638. Min Act: -0.1698

Data Index: 1435827 (Python Code)

Max Activating Token Index: 889

Text #11

Max Range: 5.5950. Min Range: -5.5950

Max Act: 5.0950. Min Act: -0.1185

Data Index: 1560703 (Python Code)

Max Activating Token Index: 493

Text #12

Max Range: 5.5950. Min Range: -5.5950

Max Act: 5.0638. Min Act: -0.1699

Data Index: 1517868 (Python Code)

Max Activating Token Index: 979

Text #13

Max Range: 5.5950. Min Range: -5.5950

Max Act: 5.0638. Min Act: -0.1700

Data Index: 1544098 (Python Code)

Max Activating Token Index: 913

Text #14

Max Range: 5.5950. Min Range: -5.5950

Max Act: 5.0638. Min Act: -0.1700

Data Index: 1001187 (C4 (Web Text))

Max Activating Token Index: 180

Text #15

Max Range: 5.5950. Min Range: -5.5950

Max Act: 4.9700. Min Act: -0.1700

Data Index: 1517427 (Python Code)

Max Activating Token Index: 532

Text #16

Max Range: 5.5950. Min Range: -5.5950

Max Act: 5.0013. Min Act: -0.1699

Data Index: 1508134 (Python Code)

Max Activating Token Index: 469

Text #17

Max Range: 5.5950. Min Range: -5.5950

Max Act: 5.0638. Min Act: -0.1698

Data Index: 280503 (C4 (Web Text))

Max Activating Token Index: 883

Text #18

Max Range: 5.5950. Min Range: -5.5950

Max Act: 4.9700. Min Act: -0.1700

Data Index: 1347468 (C4 (Web Text))

Max Activating Token Index: 486

Text #19

Max Range: 5.5950. Min Range: -5.5950

Max Act: 5.0325. Min Act: -0.1700

Data Index: 586348 (C4 (Web Text))

Max Activating Token Index: 452