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

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

Neuron 3830 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: 2.4745. Min Range: -2.4745

Max Act: 2.4745. Min Act: -0.0000

Data Index: 1526783 (Python Code)

Max Activating Token Index: 916

Text #1

Max Range: 2.4745. Min Range: -2.4745

Max Act: 2.3858. Min Act: -0.0000

Data Index: 1099861 (C4 (Web Text))

Max Activating Token Index: 22

Text #2

Max Range: 2.4745. Min Range: -2.4745

Max Act: 1.9478. Min Act: -0.0000

Data Index: 620448 (C4 (Web Text))

Max Activating Token Index: 710

Text #3

Max Range: 2.4745. Min Range: -2.4745

Max Act: 2.0401. Min Act: -0.0000

Data Index: 257487 (C4 (Web Text))

Max Activating Token Index: 396

Text #4

Max Range: 2.4745. Min Range: -2.4745

Max Act: 1.9306. Min Act: -0.0000

Data Index: 1441960 (Python Code)

Max Activating Token Index: 877

Text #5

Max Range: 2.4745. Min Range: -2.4745

Max Act: 1.7095. Min Act: -0.0000

Data Index: 1600301 (Python Code)

Max Activating Token Index: 84

Text #6

Max Range: 2.4745. Min Range: -2.4745

Max Act: 1.7845. Min Act: -0.0000

Data Index: 46493 (C4 (Web Text))

Max Activating Token Index: 296

Text #7

Max Range: 2.4745. Min Range: -2.4745

Max Act: 1.8703. Min Act: -0.0000

Data Index: 57463 (C4 (Web Text))

Max Activating Token Index: 158

Text #8

Max Range: 2.4745. Min Range: -2.4745

Max Act: 1.8081. Min Act: -0.0000

Data Index: 1640005 (Python Code)

Max Activating Token Index: 686

Text #9

Max Range: 2.4745. Min Range: -2.4745

Max Act: 1.8213. Min Act: -0.0000

Data Index: 1346975 (C4 (Web Text))

Max Activating Token Index: 93

Text #10

Max Range: 2.4745. Min Range: -2.4745

Max Act: 1.7208. Min Act: -0.0000

Data Index: 463595 (C4 (Web Text))

Max Activating Token Index: 389

Text #11

Max Range: 2.4745. Min Range: -2.4745

Max Act: 1.7104. Min Act: -0.0000

Data Index: 781450 (C4 (Web Text))

Max Activating Token Index: 486

Text #12

Max Range: 2.4745. Min Range: -2.4745

Max Act: 1.8947. Min Act: -0.0000

Data Index: 1399261 (Python Code)

Max Activating Token Index: 375

Text #13

Max Range: 2.4745. Min Range: -2.4745

Max Act: 1.8742. Min Act: -0.0000

Data Index: 1593822 (Python Code)

Max Activating Token Index: 809

Text #14

Max Range: 2.4745. Min Range: -2.4745

Max Act: 1.7114. Min Act: -0.0000

Data Index: 1387035 (Python Code)

Max Activating Token Index: 544

Text #15

Max Range: 2.4745. Min Range: -2.4745

Max Act: 1.6713. Min Act: -0.0000

Data Index: 1527144 (Python Code)

Max Activating Token Index: 773

Text #16

Max Range: 2.4745. Min Range: -2.4745

Max Act: 1.5617. Min Act: -0.0000

Data Index: 449991 (C4 (Web Text))

Max Activating Token Index: 493

Text #17

Max Range: 2.4745. Min Range: -2.4745

Max Act: 1.6593. Min Act: -0.0000

Data Index: 1479776 (Python Code)

Max Activating Token Index: 704

Text #18

Max Range: 2.4745. Min Range: -2.4745

Max Act: 1.6545. Min Act: -0.0000

Data Index: 1480417 (Python Code)

Max Activating Token Index: 460

Text #19

Max Range: 2.4745. Min Range: -2.4745

Max Act: 1.8388. Min Act: -0.0000

Data Index: 964731 (C4 (Web Text))

Max Activating Token Index: 652