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

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

Neuron 4722 in Layer 7

Load this data into an Interactive Neuroscope

See Documentation here

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

Text #0

Max Range: 2.3530. Min Range: -2.3530

Max Act: 2.3350. Min Act: -0.0000

Data Index: 1468007 (Python Code)

Max Activating Token Index: 942

Text #1

Max Range: 2.3530. Min Range: -2.3530

Max Act: 2.3530. Min Act: -0.0000

Data Index: 1645516 (Python Code)

Max Activating Token Index: 810

Text #2

Max Range: 2.3530. Min Range: -2.3530

Max Act: 1.9181. Min Act: -0.0000

Data Index: 1432611 (Python Code)

Max Activating Token Index: 680

Text #3

Max Range: 2.3530. Min Range: -2.3530

Max Act: 2.0093. Min Act: -0.0000

Data Index: 168690 (C4 (Web Text))

Max Activating Token Index: 527

Text #4

Max Range: 2.3530. Min Range: -2.3530

Max Act: 1.7546. Min Act: -0.0000

Data Index: 1637997 (Python Code)

Max Activating Token Index: 684

Text #5

Max Range: 2.3530. Min Range: -2.3530

Max Act: 2.1550. Min Act: -0.0000

Data Index: 1408770 (Python Code)

Max Activating Token Index: 887

Text #6

Max Range: 2.3530. Min Range: -2.3530

Max Act: 1.8071. Min Act: -0.0000

Data Index: 752281 (C4 (Web Text))

Max Activating Token Index: 232

Text #7

Max Range: 2.3530. Min Range: -2.3530

Max Act: 1.8989. Min Act: -0.0000

Data Index: 975326 (C4 (Web Text))

Max Activating Token Index: 165

Text #8

Max Range: 2.3530. Min Range: -2.3530

Max Act: 2.0446. Min Act: -0.0000

Data Index: 1500443 (Python Code)

Max Activating Token Index: 403

Text #9

Max Range: 2.3530. Min Range: -2.3530

Max Act: 1.9418. Min Act: -0.0000

Data Index: 1459670 (Python Code)

Max Activating Token Index: 848

Text #10

Max Range: 2.3530. Min Range: -2.3530

Max Act: 1.8345. Min Act: -0.0000

Data Index: 347028 (C4 (Web Text))

Max Activating Token Index: 336

Text #11

Max Range: 2.3530. Min Range: -2.3530

Max Act: 1.7501. Min Act: -0.0000

Data Index: 1515250 (Python Code)

Max Activating Token Index: 841

Text #12

Max Range: 2.3530. Min Range: -2.3530

Max Act: 1.7163. Min Act: -0.0000

Data Index: 296075 (C4 (Web Text))

Max Activating Token Index: 703

Text #13

Max Range: 2.3530. Min Range: -2.3530

Max Act: 1.9016. Min Act: -0.0000

Data Index: 1386420 (Python Code)

Max Activating Token Index: 402

Text #14

Max Range: 2.3530. Min Range: -2.3530

Max Act: 1.6864. Min Act: -0.0000

Data Index: 1459151 (Python Code)

Max Activating Token Index: 624

Text #15

Max Range: 2.3530. Min Range: -2.3530

Max Act: 1.7804. Min Act: -0.0000

Data Index: 142432 (C4 (Web Text))

Max Activating Token Index: 1011

Text #16

Max Range: 2.3530. Min Range: -2.3530

Max Act: 1.6771. Min Act: -0.0000

Data Index: 1515282 (Python Code)

Max Activating Token Index: 695

Text #17

Max Range: 2.3530. Min Range: -2.3530

Max Act: 1.6379. Min Act: -0.0000

Data Index: 494720 (C4 (Web Text))

Max Activating Token Index: 561

Text #18

Max Range: 2.3530. Min Range: -2.3530

Max Act: 1.6232. Min Act: -0.0000

Data Index: 1532114 (Python Code)

Max Activating Token Index: 179

Text #19

Max Range: 2.3530. Min Range: -2.3530

Max Act: 1.6856. Min Act: -0.0000

Data Index: 1454060 (Python Code)

Max Activating Token Index: 644