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

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

Neuron 337 in Layer 2

Load this data into an Interactive Neuroscope

See Documentation here

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

Text #0

Max Range: 3.6710. Min Range: -3.6710

Max Act: 3.6710. Min Act: -0.1700

Data Index: 1100833 (C4 (Web Text))

Max Activating Token Index: 1023

Text #1

Max Range: 3.6710. Min Range: -3.6710

Max Act: 3.6710. Min Act: -0.1700

Data Index: 1628939 (Python Code)

Max Activating Token Index: 514

Text #2

Max Range: 3.6710. Min Range: -3.6710

Max Act: 3.6710. Min Act: -0.1700

Data Index: 1195830 (C4 (Web Text))

Max Activating Token Index: 778

Text #3

Max Range: 3.6710. Min Range: -3.6710

Max Act: 3.6240. Min Act: -0.1700

Data Index: 1386128 (Python Code)

Max Activating Token Index: 451

Text #4

Max Range: 3.6710. Min Range: -3.6710

Max Act: 3.6084. Min Act: -0.1700

Data Index: 1451148 (Python Code)

Max Activating Token Index: 746

Text #5

Max Range: 3.6710. Min Range: -3.6710

Max Act: 3.5614. Min Act: -0.1700

Data Index: 1639986 (Python Code)

Max Activating Token Index: 339

Text #6

Max Range: 3.6710. Min Range: -3.6710

Max Act: 3.5771. Min Act: -0.1700

Data Index: 1446641 (Python Code)

Max Activating Token Index: 713

Text #7

Max Range: 3.6710. Min Range: -3.6710

Max Act: 3.5771. Min Act: -0.1700

Data Index: 1633105 (Python Code)

Max Activating Token Index: 422

Text #8

Max Range: 3.6710. Min Range: -3.6710

Max Act: 3.5457. Min Act: -0.1700

Data Index: 713640 (C4 (Web Text))

Max Activating Token Index: 1003

Text #9

Max Range: 3.6710. Min Range: -3.6710

Max Act: 3.5614. Min Act: -0.1700

Data Index: 321378 (C4 (Web Text))

Max Activating Token Index: 113

Text #10

Max Range: 3.6710. Min Range: -3.6710

Max Act: 3.5301. Min Act: -0.1700

Data Index: 196210 (C4 (Web Text))

Max Activating Token Index: 402

Text #11

Max Range: 3.6710. Min Range: -3.6710

Max Act: 3.5301. Min Act: -0.1700

Data Index: 253120 (C4 (Web Text))

Max Activating Token Index: 437

Text #12

Max Range: 3.6710. Min Range: -3.6710

Max Act: 3.5301. Min Act: -0.1700

Data Index: 1440159 (Python Code)

Max Activating Token Index: 514

Text #13

Max Range: 3.6710. Min Range: -3.6710

Max Act: 3.5457. Min Act: -0.1700

Data Index: 1567680 (Python Code)

Max Activating Token Index: 819

Text #14

Max Range: 3.6710. Min Range: -3.6710

Max Act: 3.5457. Min Act: -0.1700

Data Index: 29838 (C4 (Web Text))

Max Activating Token Index: 884

Text #15

Max Range: 3.6710. Min Range: -3.6710

Max Act: 3.5144. Min Act: -0.1700

Data Index: 1231937 (C4 (Web Text))

Max Activating Token Index: 355

Text #16

Max Range: 3.6710. Min Range: -3.6710

Max Act: 3.5301. Min Act: -0.1700

Data Index: 1075520 (C4 (Web Text))

Max Activating Token Index: 624

Text #17

Max Range: 3.6710. Min Range: -3.6710

Max Act: 3.4987. Min Act: -0.1700

Data Index: 1453110 (Python Code)

Max Activating Token Index: 223

Text #18

Max Range: 3.6710. Min Range: -3.6710

Max Act: 3.5301. Min Act: -0.1700

Data Index: 1363204 (Python Code)

Max Activating Token Index: 201

Text #19

Max Range: 3.6710. Min Range: -3.6710

Max Act: 3.5144. Min Act: -0.1700

Data Index: 290068 (C4 (Web Text))

Max Activating Token Index: 627