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

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

Neuron 5221 in Layer 6

Load this data into an Interactive Neuroscope

See Documentation here

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

Text #0

Max Range: 4.4758. Min Range: -4.4758

Max Act: 4.4758. Min Act: -0.0000

Data Index: 547689 (C4 (Web Text))

Max Activating Token Index: 572

Text #1

Max Range: 4.4758. Min Range: -4.4758

Max Act: 4.0575. Min Act: -0.0000

Data Index: 1415134 (Python Code)

Max Activating Token Index: 380

Text #2

Max Range: 4.4758. Min Range: -4.4758

Max Act: 4.0854. Min Act: -0.0000

Data Index: 1369040 (Python Code)

Max Activating Token Index: 947

Text #3

Max Range: 4.4758. Min Range: -4.4758

Max Act: 3.9957. Min Act: -0.0000

Data Index: 1460650 (Python Code)

Max Activating Token Index: 950

Text #4

Max Range: 4.4758. Min Range: -4.4758

Max Act: 3.8164. Min Act: -0.0000

Data Index: 1192285 (C4 (Web Text))

Max Activating Token Index: 68

Text #5

Max Range: 4.4758. Min Range: -4.4758

Max Act: 3.4950. Min Act: -0.0000

Data Index: 1412994 (Python Code)

Max Activating Token Index: 460

Text #6

Max Range: 4.4758. Min Range: -4.4758

Max Act: 3.3239. Min Act: -0.0000

Data Index: 672190 (C4 (Web Text))

Max Activating Token Index: 629

Text #7

Max Range: 4.4758. Min Range: -4.4758

Max Act: 3.1332. Min Act: -0.0000

Data Index: 1619071 (Python Code)

Max Activating Token Index: 902

Text #8

Max Range: 4.4758. Min Range: -4.4758

Max Act: 3.0421. Min Act: -0.0000

Data Index: 1558701 (Python Code)

Max Activating Token Index: 861

Text #9

Max Range: 4.4758. Min Range: -4.4758

Max Act: 3.1518. Min Act: -0.0000

Data Index: 1508505 (Python Code)

Max Activating Token Index: 35

Text #10

Max Range: 4.4758. Min Range: -4.4758

Max Act: 3.2808. Min Act: -0.0000

Data Index: 354763 (C4 (Web Text))

Max Activating Token Index: 975

Text #11

Max Range: 4.4758. Min Range: -4.4758

Max Act: 2.9636. Min Act: -0.0000

Data Index: 1605161 (Python Code)

Max Activating Token Index: 824

Text #12

Max Range: 4.4758. Min Range: -4.4758

Max Act: 2.9325. Min Act: -0.0000

Data Index: 1654513 (Python Code)

Max Activating Token Index: 1019

Text #13

Max Range: 4.4758. Min Range: -4.4758

Max Act: 3.0220. Min Act: -0.0000

Data Index: 1615679 (Python Code)

Max Activating Token Index: 142

Text #14

Max Range: 4.4758. Min Range: -4.4758

Max Act: 2.9427. Min Act: -0.0000

Data Index: 1407169 (Python Code)

Max Activating Token Index: 306

Text #15

Max Range: 4.4758. Min Range: -4.4758

Max Act: 2.9505. Min Act: -0.0000

Data Index: 1421014 (Python Code)

Max Activating Token Index: 483

Text #16

Max Range: 4.4758. Min Range: -4.4758

Max Act: 2.9254. Min Act: -0.0000

Data Index: 1556164 (Python Code)

Max Activating Token Index: 596

Text #17

Max Range: 4.4758. Min Range: -4.4758

Max Act: 3.0311. Min Act: -0.0000

Data Index: 138126 (C4 (Web Text))

Max Activating Token Index: 936

Text #18

Max Range: 4.4758. Min Range: -4.4758

Max Act: 2.7468. Min Act: -0.0000

Data Index: 1379630 (Python Code)

Max Activating Token Index: 971

Text #19

Max Range: 4.4758. Min Range: -4.4758

Max Act: 2.9974. Min Act: -0.0000

Data Index: 432079 (C4 (Web Text))

Max Activating Token Index: 310