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

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

Neuron 3939 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: 3.8485. Min Range: -3.8485

Max Act: 3.8485. Min Act: -0.0000

Data Index: 1351721 (C4 (Web Text))

Max Activating Token Index: 861

Text #1

Max Range: 3.8485. Min Range: -3.8485

Max Act: 2.5997. Min Act: -0.0000

Data Index: 1183938 (C4 (Web Text))

Max Activating Token Index: 958

Text #2

Max Range: 3.8485. Min Range: -3.8485

Max Act: 2.4717. Min Act: -0.0000

Data Index: 1430238 (Python Code)

Max Activating Token Index: 381

Text #3

Max Range: 3.8485. Min Range: -3.8485

Max Act: 2.3900. Min Act: -0.0000

Data Index: 1384214 (Python Code)

Max Activating Token Index: 431

Text #4

Max Range: 3.8485. Min Range: -3.8485

Max Act: 2.2885. Min Act: -0.0000

Data Index: 295052 (C4 (Web Text))

Max Activating Token Index: 375

Text #5

Max Range: 3.8485. Min Range: -3.8485

Max Act: 2.2916. Min Act: -0.0000

Data Index: 1212022 (C4 (Web Text))

Max Activating Token Index: 863

Text #6

Max Range: 3.8485. Min Range: -3.8485

Max Act: 2.3645. Min Act: -0.0000

Data Index: 1361664 (Python Code)

Max Activating Token Index: 690

Text #7

Max Range: 3.8485. Min Range: -3.8485

Max Act: 2.3288. Min Act: -0.0000

Data Index: 585872 (C4 (Web Text))

Max Activating Token Index: 453

Text #8

Max Range: 3.8485. Min Range: -3.8485

Max Act: 1.8869. Min Act: -0.0000

Data Index: 30630 (C4 (Web Text))

Max Activating Token Index: 148

Text #9

Max Range: 3.8485. Min Range: -3.8485

Max Act: 2.3370. Min Act: -0.0000

Data Index: 978154 (C4 (Web Text))

Max Activating Token Index: 1017

Text #10

Max Range: 3.8485. Min Range: -3.8485

Max Act: 2.1133. Min Act: -0.0000

Data Index: 1584760 (Python Code)

Max Activating Token Index: 501

Text #11

Max Range: 3.8485. Min Range: -3.8485

Max Act: 2.4783. Min Act: -0.0000

Data Index: 876962 (C4 (Web Text))

Max Activating Token Index: 469

Text #12

Max Range: 3.8485. Min Range: -3.8485

Max Act: 2.1178. Min Act: -0.0000

Data Index: 577951 (C4 (Web Text))

Max Activating Token Index: 139

Text #13

Max Range: 3.8485. Min Range: -3.8485

Max Act: 2.3465. Min Act: -0.0000

Data Index: 1149951 (C4 (Web Text))

Max Activating Token Index: 234

Text #14

Max Range: 3.8485. Min Range: -3.8485

Max Act: 2.3221. Min Act: -0.0000

Data Index: 614961 (C4 (Web Text))

Max Activating Token Index: 431

Text #15

Max Range: 3.8485. Min Range: -3.8485

Max Act: 2.2021. Min Act: -0.0000

Data Index: 1412382 (Python Code)

Max Activating Token Index: 147

Text #16

Max Range: 3.8485. Min Range: -3.8485

Max Act: 2.0749. Min Act: -0.0000

Data Index: 202189 (C4 (Web Text))

Max Activating Token Index: 461

Text #17

Max Range: 3.8485. Min Range: -3.8485

Max Act: 1.9515. Min Act: -0.0000

Data Index: 1538033 (Python Code)

Max Activating Token Index: 405

Text #18

Max Range: 3.8485. Min Range: -3.8485

Max Act: 2.1629. Min Act: -0.0000

Data Index: 380028 (C4 (Web Text))

Max Activating Token Index: 555

Text #19

Max Range: 3.8485. Min Range: -3.8485

Max Act: 2.1594. Min Act: -0.0000

Data Index: 1655650 (Python Code)

Max Activating Token Index: 510