Model: SoLU Model: 12 Layers, 6144 Neurons per Layer

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

Neuron 4536 in Layer 10

Load this data into an Interactive Neuroscope

See Documentation here

Transformer Lens Loading: HookedTransformer.from_pretrained('solu-12l')

Text #0

Max Range: 2.8471. Min Range: -2.8471

Max Act: 2.8471. Min Act: -0.0000

Data Index: 829438 (C4 (Web Text))

Max Activating Token Index: 395

Text #1

Max Range: 2.8471. Min Range: -2.8471

Max Act: 2.6947. Min Act: -0.0000

Data Index: 94075 (C4 (Web Text))

Max Activating Token Index: 450

Text #2

Max Range: 2.8471. Min Range: -2.8471

Max Act: 1.9983. Min Act: -0.0000

Data Index: 1250613 (C4 (Web Text))

Max Activating Token Index: 638

Text #3

Max Range: 2.8471. Min Range: -2.8471

Max Act: 1.8936. Min Act: -0.0000

Data Index: 817325 (C4 (Web Text))

Max Activating Token Index: 470

Text #4

Max Range: 2.8471. Min Range: -2.8471

Max Act: 2.0785. Min Act: 0.0001

Data Index: 1235317 (C4 (Web Text))

Max Activating Token Index: 327

Text #5

Max Range: 2.8471. Min Range: -2.8471

Max Act: 1.5839. Min Act: -0.0000

Data Index: 1286186 (C4 (Web Text))

Max Activating Token Index: 76

Text #6

Max Range: 2.8471. Min Range: -2.8471

Max Act: 1.6904. Min Act: -0.0000

Data Index: 1273828 (C4 (Web Text))

Max Activating Token Index: 800

Text #7

Max Range: 2.8471. Min Range: -2.8471

Max Act: 1.5418. Min Act: 0.0000

Data Index: 125429 (C4 (Web Text))

Max Activating Token Index: 45

Text #8

Max Range: 2.8471. Min Range: -2.8471

Max Act: 1.4616. Min Act: -0.0000

Data Index: 739602 (C4 (Web Text))

Max Activating Token Index: 1020

Text #9

Max Range: 2.8471. Min Range: -2.8471

Max Act: 1.5433. Min Act: -0.0000

Data Index: 691848 (C4 (Web Text))

Max Activating Token Index: 744

Text #10

Max Range: 2.8471. Min Range: -2.8471

Max Act: 1.5249. Min Act: -0.0000

Data Index: 220139 (C4 (Web Text))

Max Activating Token Index: 177

Text #11

Max Range: 2.8471. Min Range: -2.8471

Max Act: 1.4269. Min Act: 0.0000

Data Index: 1251455 (C4 (Web Text))

Max Activating Token Index: 447

Text #12

Max Range: 2.8471. Min Range: -2.8471

Max Act: 1.3801. Min Act: -0.0000

Data Index: 1335652 (C4 (Web Text))

Max Activating Token Index: 91

Text #13

Max Range: 2.8471. Min Range: -2.8471

Max Act: 1.3870. Min Act: -0.0000

Data Index: 1561015 (Python Code)

Max Activating Token Index: 946

Text #14

Max Range: 2.8471. Min Range: -2.8471

Max Act: 1.3578. Min Act: -0.0000

Data Index: 1515108 (Python Code)

Max Activating Token Index: 836

Text #15

Max Range: 2.8471. Min Range: -2.8471

Max Act: 1.1215. Min Act: 0.0000

Data Index: 1358328 (C4 (Web Text))

Max Activating Token Index: 231

Text #16

Max Range: 2.8471. Min Range: -2.8471

Max Act: 1.3524. Min Act: 0.0000

Data Index: 89995 (C4 (Web Text))

Max Activating Token Index: 705

Text #17

Max Range: 2.8471. Min Range: -2.8471

Max Act: 1.3110. Min Act: -0.0000

Data Index: 1128043 (C4 (Web Text))

Max Activating Token Index: 638

Text #18

Max Range: 2.8471. Min Range: -2.8471

Max Act: 1.3637. Min Act: -0.0000

Data Index: 77306 (C4 (Web Text))

Max Activating Token Index: 834

Text #19

Max Range: 2.8471. Min Range: -2.8471

Max Act: 1.2678. Min Act: -0.0000

Data Index: 1020719 (C4 (Web Text))

Max Activating Token Index: 456