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

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

Neuron 3945 in Layer 5

Load this data into an Interactive Neuroscope

See Documentation here

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

Text #0

Max Range: 0.1095. Min Range: -0.1095

Max Act: 0.1095. Min Act: -0.0001

Data Index: 1042697 (C4 (Web Text))

Max Activating Token Index: 1001

Text #1

Max Range: 0.1095. Min Range: -0.1095

Max Act: 0.1050. Min Act: -0.0001

Data Index: 657567 (C4 (Web Text))

Max Activating Token Index: 207

Text #2

Max Range: 0.1095. Min Range: -0.1095

Max Act: 0.1018. Min Act: -0.0001

Data Index: 220520 (C4 (Web Text))

Max Activating Token Index: 251

Text #3

Max Range: 0.1095. Min Range: -0.1095

Max Act: 0.0842. Min Act: -0.0001

Data Index: 1552252 (Python Code)

Max Activating Token Index: 895

Text #4

Max Range: 0.1095. Min Range: -0.1095

Max Act: 0.0901. Min Act: -0.0001

Data Index: 851797 (C4 (Web Text))

Max Activating Token Index: 860

Text #5

Max Range: 0.1095. Min Range: -0.1095

Max Act: 0.0916. Min Act: -0.0001

Data Index: 1465845 (Python Code)

Max Activating Token Index: 599

Text #6

Max Range: 0.1095. Min Range: -0.1095

Max Act: 0.0896. Min Act: -0.0001

Data Index: 228450 (C4 (Web Text))

Max Activating Token Index: 421

Text #7

Max Range: 0.1095. Min Range: -0.1095

Max Act: 0.0795. Min Act: -0.0001

Data Index: 874953 (C4 (Web Text))

Max Activating Token Index: 15

Text #8

Max Range: 0.1095. Min Range: -0.1095

Max Act: 0.0882. Min Act: -0.0001

Data Index: 19414 (C4 (Web Text))

Max Activating Token Index: 679

Text #9

Max Range: 0.1095. Min Range: -0.1095

Max Act: 0.0782. Min Act: -0.0001

Data Index: 1149920 (C4 (Web Text))

Max Activating Token Index: 4

Text #10

Max Range: 0.1095. Min Range: -0.1095

Max Act: 0.0808. Min Act: -0.0001

Data Index: 1451624 (Python Code)

Max Activating Token Index: 559

Text #11

Max Range: 0.1095. Min Range: -0.1095

Max Act: 0.0819. Min Act: -0.0001

Data Index: 1540260 (Python Code)

Max Activating Token Index: 585

Text #12

Max Range: 0.1095. Min Range: -0.1095

Max Act: 0.0769. Min Act: -0.0001

Data Index: 507568 (C4 (Web Text))

Max Activating Token Index: 804

Text #13

Max Range: 0.1095. Min Range: -0.1095

Max Act: 0.0700. Min Act: -0.0001

Data Index: 340581 (C4 (Web Text))

Max Activating Token Index: 274

Text #14

Max Range: 0.1095. Min Range: -0.1095

Max Act: 0.0760. Min Act: -0.0001

Data Index: 596434 (C4 (Web Text))

Max Activating Token Index: 976

Text #15

Max Range: 0.1095. Min Range: -0.1095

Max Act: 0.0759. Min Act: -0.0001

Data Index: 387934 (C4 (Web Text))

Max Activating Token Index: 227

Text #16

Max Range: 0.1095. Min Range: -0.1095

Max Act: 0.0688. Min Act: -0.0001

Data Index: 448923 (C4 (Web Text))

Max Activating Token Index: 278

Text #17

Max Range: 0.1095. Min Range: -0.1095

Max Act: 0.0708. Min Act: -0.0001

Data Index: 92674 (C4 (Web Text))

Max Activating Token Index: 813

Text #18

Max Range: 0.1095. Min Range: -0.1095

Max Act: 0.0725. Min Act: -0.0001

Data Index: 880309 (C4 (Web Text))

Max Activating Token Index: 18

Text #19

Max Range: 0.1095. Min Range: -0.1095

Max Act: 0.0677. Min Act: -0.0001

Data Index: 1454515 (Python Code)

Max Activating Token Index: 396