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

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

Neuron 525 in Layer 1

Load this data into an Interactive Neuroscope

See Documentation here

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

Text #0

Max Range: 3.3740. Min Range: -3.3740

Max Act: 3.3740. Min Act: -0.1700

Data Index: 797531 (C4 (Web Text))

Max Activating Token Index: 967

Text #1

Max Range: 3.3740. Min Range: -3.3740

Max Act: 3.3426. Min Act: -0.1700

Data Index: 321307 (C4 (Web Text))

Max Activating Token Index: 832

Text #2

Max Range: 3.3740. Min Range: -3.3740

Max Act: 3.3269. Min Act: -0.1700

Data Index: 980735 (C4 (Web Text))

Max Activating Token Index: 93

Text #3

Max Range: 3.3740. Min Range: -3.3740

Max Act: 3.2641. Min Act: -0.1700

Data Index: 884724 (C4 (Web Text))

Max Activating Token Index: 206

Text #4

Max Range: 3.3740. Min Range: -3.3740

Max Act: 3.2484. Min Act: -0.1700

Data Index: 163174 (C4 (Web Text))

Max Activating Token Index: 818

Text #5

Max Range: 3.3740. Min Range: -3.3740

Max Act: 3.2641. Min Act: -0.1700

Data Index: 236299 (C4 (Web Text))

Max Activating Token Index: 12

Text #6

Max Range: 3.3740. Min Range: -3.3740

Max Act: 3.2641. Min Act: -0.1700

Data Index: 222321 (C4 (Web Text))

Max Activating Token Index: 749

Text #7

Max Range: 3.3740. Min Range: -3.3740

Max Act: 3.2484. Min Act: -0.1700

Data Index: 603784 (C4 (Web Text))

Max Activating Token Index: 126

Text #8

Max Range: 3.3740. Min Range: -3.3740

Max Act: 3.2484. Min Act: -0.1700

Data Index: 554376 (C4 (Web Text))

Max Activating Token Index: 612

Text #9

Max Range: 3.3740. Min Range: -3.3740

Max Act: 3.2641. Min Act: -0.1700

Data Index: 1266074 (C4 (Web Text))

Max Activating Token Index: 130

Text #10

Max Range: 3.3740. Min Range: -3.3740

Max Act: 3.2641. Min Act: -0.1700

Data Index: 288190 (C4 (Web Text))

Max Activating Token Index: 597

Text #11

Max Range: 3.3740. Min Range: -3.3740

Max Act: 3.2169. Min Act: -0.1700

Data Index: 1282678 (C4 (Web Text))

Max Activating Token Index: 560

Text #12

Max Range: 3.3740. Min Range: -3.3740

Max Act: 3.2012. Min Act: -0.1700

Data Index: 729740 (C4 (Web Text))

Max Activating Token Index: 91

Text #13

Max Range: 3.3740. Min Range: -3.3740

Max Act: 3.1855. Min Act: -0.1700

Data Index: 1436591 (Python Code)

Max Activating Token Index: 735

Text #14

Max Range: 3.3740. Min Range: -3.3740

Max Act: 3.2169. Min Act: -0.1700

Data Index: 1337682 (C4 (Web Text))

Max Activating Token Index: 988

Text #15

Max Range: 3.3740. Min Range: -3.3740

Max Act: 3.2484. Min Act: -0.1700

Data Index: 1228674 (C4 (Web Text))

Max Activating Token Index: 807

Text #16

Max Range: 3.3740. Min Range: -3.3740

Max Act: 3.1855. Min Act: -0.1700

Data Index: 334031 (C4 (Web Text))

Max Activating Token Index: 215

Text #17

Max Range: 3.3740. Min Range: -3.3740

Max Act: 3.1855. Min Act: -0.1700

Data Index: 1224406 (C4 (Web Text))

Max Activating Token Index: 584

Text #18

Max Range: 3.3740. Min Range: -3.3740

Max Act: 3.1855. Min Act: -0.1700

Data Index: 866764 (C4 (Web Text))

Max Activating Token Index: 795

Text #19

Max Range: 3.3740. Min Range: -3.3740

Max Act: 3.1855. Min Act: -0.1700

Data Index: 231705 (C4 (Web Text))

Max Activating Token Index: 579