Model: SoLU Model: 10 Layers, 5120 Neurons per Layer

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

Neuron 3814 in Layer 7

Load this data into an Interactive Neuroscope

See Documentation here

Transformer Lens Loading: HookedTransformer.from_pretrained('solu-10l')

Text #0

Max Range: 3.5607. Min Range: -3.5607

Max Act: 3.5607. Min Act: -0.0000

Data Index: 1003703 (C4 (Web Text))

Max Activating Token Index: 115

Text #1

Max Range: 3.5607. Min Range: -3.5607

Max Act: 2.8825. Min Act: -0.0000

Data Index: 1001937 (C4 (Web Text))

Max Activating Token Index: 83

Text #2

Max Range: 3.5607. Min Range: -3.5607

Max Act: 2.8097. Min Act: -0.0000

Data Index: 182908 (C4 (Web Text))

Max Activating Token Index: 311

Text #3

Max Range: 3.5607. Min Range: -3.5607

Max Act: 2.9164. Min Act: -0.0000

Data Index: 263932 (C4 (Web Text))

Max Activating Token Index: 123

Text #4

Max Range: 3.5607. Min Range: -3.5607

Max Act: 2.6682. Min Act: -0.0000

Data Index: 135303 (C4 (Web Text))

Max Activating Token Index: 94

Text #5

Max Range: 3.5607. Min Range: -3.5607

Max Act: 2.7564. Min Act: -0.0000

Data Index: 1330714 (C4 (Web Text))

Max Activating Token Index: 186

Text #6

Max Range: 3.5607. Min Range: -3.5607

Max Act: 2.8065. Min Act: -0.0000

Data Index: 1487242 (Python Code)

Max Activating Token Index: 191

Text #7

Max Range: 3.5607. Min Range: -3.5607

Max Act: 2.5388. Min Act: -0.0000

Data Index: 1623782 (Python Code)

Max Activating Token Index: 586

Text #8

Max Range: 3.5607. Min Range: -3.5607

Max Act: 2.6771. Min Act: -0.0000

Data Index: 1557685 (Python Code)

Max Activating Token Index: 234

Text #9

Max Range: 3.5607. Min Range: -3.5607

Max Act: 2.5196. Min Act: -0.0000

Data Index: 1276892 (C4 (Web Text))

Max Activating Token Index: 473

Text #10

Max Range: 3.5607. Min Range: -3.5607

Max Act: 2.5216. Min Act: -0.0000

Data Index: 459579 (C4 (Web Text))

Max Activating Token Index: 156

Text #11

Max Range: 3.5607. Min Range: -3.5607

Max Act: 2.5022. Min Act: -0.0000

Data Index: 1428079 (Python Code)

Max Activating Token Index: 222

Text #12

Max Range: 3.5607. Min Range: -3.5607

Max Act: 2.5794. Min Act: -0.0000

Data Index: 1391642 (Python Code)

Max Activating Token Index: 1006

Text #13

Max Range: 3.5607. Min Range: -3.5607

Max Act: 2.5318. Min Act: -0.0000

Data Index: 1463105 (Python Code)

Max Activating Token Index: 176

Text #14

Max Range: 3.5607. Min Range: -3.5607

Max Act: 2.4767. Min Act: -0.0000

Data Index: 1583004 (Python Code)

Max Activating Token Index: 242

Text #15

Max Range: 3.5607. Min Range: -3.5607

Max Act: 2.3662. Min Act: -0.0000

Data Index: 1419742 (Python Code)

Max Activating Token Index: 587

Text #16

Max Range: 3.5607. Min Range: -3.5607

Max Act: 2.4828. Min Act: -0.0000

Data Index: 1146867 (C4 (Web Text))

Max Activating Token Index: 142

Text #17

Max Range: 3.5607. Min Range: -3.5607

Max Act: 2.4763. Min Act: -0.0000

Data Index: 482242 (C4 (Web Text))

Max Activating Token Index: 939

Text #18

Max Range: 3.5607. Min Range: -3.5607

Max Act: 2.3423. Min Act: -0.0000

Data Index: 1182872 (C4 (Web Text))

Max Activating Token Index: 904

Text #19

Max Range: 3.5607. Min Range: -3.5607

Max Act: 2.4484. Min Act: -0.0000

Data Index: 1648908 (Python Code)

Max Activating Token Index: 935