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

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

Neuron 1159 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: 5.3723. Min Range: -5.3723

Max Act: 5.3723. Min Act: -0.1700

Data Index: 1233907 (C4 (Web Text))

Max Activating Token Index: 115

Text #1

Max Range: 5.3723. Min Range: -5.3723

Max Act: 5.3410. Min Act: -0.1700

Data Index: 823355 (C4 (Web Text))

Max Activating Token Index: 960

Text #2

Max Range: 5.3723. Min Range: -5.3723

Max Act: 5.3723. Min Act: -0.1700

Data Index: 633789 (C4 (Web Text))

Max Activating Token Index: 347

Text #3

Max Range: 5.3723. Min Range: -5.3723

Max Act: 5.3410. Min Act: -0.1700

Data Index: 833678 (C4 (Web Text))

Max Activating Token Index: 849

Text #4

Max Range: 5.3723. Min Range: -5.3723

Max Act: 5.3098. Min Act: -0.1700

Data Index: 265140 (C4 (Web Text))

Max Activating Token Index: 992

Text #5

Max Range: 5.3723. Min Range: -5.3723

Max Act: 5.2785. Min Act: -0.1700

Data Index: 493579 (C4 (Web Text))

Max Activating Token Index: 64

Text #6

Max Range: 5.3723. Min Range: -5.3723

Max Act: 5.2785. Min Act: -0.1700

Data Index: 629180 (C4 (Web Text))

Max Activating Token Index: 126

Text #7

Max Range: 5.3723. Min Range: -5.3723

Max Act: 5.2160. Min Act: -0.1700

Data Index: 541056 (C4 (Web Text))

Max Activating Token Index: 799

Text #8

Max Range: 5.3723. Min Range: -5.3723

Max Act: 5.1848. Min Act: -0.1700

Data Index: 1522230 (Python Code)

Max Activating Token Index: 82

Text #9

Max Range: 5.3723. Min Range: -5.3723

Max Act: 5.1535. Min Act: -0.1700

Data Index: 270618 (C4 (Web Text))

Max Activating Token Index: 410

Text #10

Max Range: 5.3723. Min Range: -5.3723

Max Act: 5.1535. Min Act: -0.1700

Data Index: 1559739 (Python Code)

Max Activating Token Index: 977

Text #11

Max Range: 5.3723. Min Range: -5.3723

Max Act: 5.1223. Min Act: -0.1700

Data Index: 983251 (C4 (Web Text))

Max Activating Token Index: 864

Text #12

Max Range: 5.3723. Min Range: -5.3723

Max Act: 5.1535. Min Act: -0.1700

Data Index: 1121373 (C4 (Web Text))

Max Activating Token Index: 581

Text #13

Max Range: 5.3723. Min Range: -5.3723

Max Act: 5.1848. Min Act: -0.1700

Data Index: 467149 (C4 (Web Text))

Max Activating Token Index: 983

Text #14

Max Range: 5.3723. Min Range: -5.3723

Max Act: 5.1223. Min Act: -0.1699

Data Index: 1135362 (C4 (Web Text))

Max Activating Token Index: 1003

Text #15

Max Range: 5.3723. Min Range: -5.3723

Max Act: 5.0910. Min Act: -0.1700

Data Index: 666899 (C4 (Web Text))

Max Activating Token Index: 666

Text #16

Max Range: 5.3723. Min Range: -5.3723

Max Act: 5.1223. Min Act: -0.1700

Data Index: 147555 (C4 (Web Text))

Max Activating Token Index: 415

Text #17

Max Range: 5.3723. Min Range: -5.3723

Max Act: 5.1223. Min Act: -0.1698

Data Index: 720541 (C4 (Web Text))

Max Activating Token Index: 221

Text #18

Max Range: 5.3723. Min Range: -5.3723

Max Act: 5.0910. Min Act: -0.1700

Data Index: 11686 (C4 (Web Text))

Max Activating Token Index: 886

Text #19

Max Range: 5.3723. Min Range: -5.3723

Max Act: 5.0598. Min Act: -0.1700

Data Index: 3498 (C4 (Web Text))

Max Activating Token Index: 31