Model: GELU Model: 2 Layers, 2048 Neurons per Layer

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

Neuron 1950 in Layer 1

Load this data into an Interactive Neuroscope

See Documentation here

Transformer Lens Loading: HookedTransformer.from_pretrained('gelu-2l')

Text #0

Max Range: 5.1769. Min Range: -5.1769

Max Act: 5.1769. Min Act: -0.1700

Data Index: 766990 (C4 (Web Text))

Max Activating Token Index: 460

Text #1

Max Range: 5.1769. Min Range: -5.1769

Max Act: 5.0832. Min Act: -0.1700

Data Index: 925928 (C4 (Web Text))

Max Activating Token Index: 741

Text #2

Max Range: 5.1769. Min Range: -5.1769

Max Act: 5.1457. Min Act: -0.1700

Data Index: 484275 (C4 (Web Text))

Max Activating Token Index: 401

Text #3

Max Range: 5.1769. Min Range: -5.1769

Max Act: 5.1144. Min Act: -0.1700

Data Index: 281348 (C4 (Web Text))

Max Activating Token Index: 581

Text #4

Max Range: 5.1769. Min Range: -5.1769

Max Act: 5.0519. Min Act: -0.1700

Data Index: 554897 (C4 (Web Text))

Max Activating Token Index: 458

Text #5

Max Range: 5.1769. Min Range: -5.1769

Max Act: 5.0519. Min Act: -0.1700

Data Index: 838866 (C4 (Web Text))

Max Activating Token Index: 379

Text #6

Max Range: 5.1769. Min Range: -5.1769

Max Act: 5.0519. Min Act: -0.1700

Data Index: 1341543 (C4 (Web Text))

Max Activating Token Index: 692

Text #7

Max Range: 5.1769. Min Range: -5.1769

Max Act: 4.9269. Min Act: -0.1700

Data Index: 1633210 (Python Code)

Max Activating Token Index: 27

Text #8

Max Range: 5.1769. Min Range: -5.1769

Max Act: 4.9582. Min Act: -0.1700

Data Index: 1523117 (Python Code)

Max Activating Token Index: 7

Text #9

Max Range: 5.1769. Min Range: -5.1769

Max Act: 4.9582. Min Act: -0.1700

Data Index: 1169311 (C4 (Web Text))

Max Activating Token Index: 814

Text #10

Max Range: 5.1769. Min Range: -5.1769

Max Act: 4.9269. Min Act: -0.1700

Data Index: 1209255 (C4 (Web Text))

Max Activating Token Index: 45

Text #11

Max Range: 5.1769. Min Range: -5.1769

Max Act: 4.8644. Min Act: -0.1700

Data Index: 1538793 (Python Code)

Max Activating Token Index: 236

Text #12

Max Range: 5.1769. Min Range: -5.1769

Max Act: 4.8644. Min Act: -0.1700

Data Index: 1560731 (Python Code)

Max Activating Token Index: 220

Text #13

Max Range: 5.1769. Min Range: -5.1769

Max Act: 4.9269. Min Act: -0.1700

Data Index: 347660 (C4 (Web Text))

Max Activating Token Index: 161

Text #14

Max Range: 5.1769. Min Range: -5.1769

Max Act: 4.9269. Min Act: -0.1700

Data Index: 1346008 (C4 (Web Text))

Max Activating Token Index: 107

Text #15

Max Range: 5.1769. Min Range: -5.1769

Max Act: 4.9269. Min Act: -0.1700

Data Index: 464178 (C4 (Web Text))

Max Activating Token Index: 134

Text #16

Max Range: 5.1769. Min Range: -5.1769

Max Act: 4.9269. Min Act: -0.1700

Data Index: 364254 (C4 (Web Text))

Max Activating Token Index: 531

Text #17

Max Range: 5.1769. Min Range: -5.1769

Max Act: 4.8332. Min Act: -0.1700

Data Index: 1389551 (Python Code)

Max Activating Token Index: 132

Text #18

Max Range: 5.1769. Min Range: -5.1769

Max Act: 4.7082. Min Act: -0.1700

Data Index: 1426034 (Python Code)

Max Activating Token Index: 891

Text #19

Max Range: 5.1769. Min Range: -5.1769

Max Act: 4.8332. Min Act: -0.1700

Data Index: 42402 (C4 (Web Text))

Max Activating Token Index: 268