Model: SoLU Model: 8 Layers, 4096 Neurons per Layer

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

Neuron 3636 in Layer 7

Load this data into an Interactive Neuroscope

See Documentation here

Transformer Lens Loading: HookedTransformer.from_pretrained('solu-8l')

Text #0

Max Range: 0.4598. Min Range: -0.4598

Max Act: 0.4598. Min Act: -0.0000

Data Index: 1348731 (C4 (Web Text))

Max Activating Token Index: 422

Text #1

Max Range: 0.4598. Min Range: -0.4598

Max Act: 0.4238. Min Act: -0.0000

Data Index: 219841 (C4 (Web Text))

Max Activating Token Index: 454

Text #2

Max Range: 0.4598. Min Range: -0.4598

Max Act: 0.3589. Min Act: -0.0000

Data Index: 528423 (C4 (Web Text))

Max Activating Token Index: 144

Text #3

Max Range: 0.4598. Min Range: -0.4598

Max Act: 0.3537. Min Act: -0.0000

Data Index: 814527 (C4 (Web Text))

Max Activating Token Index: 492

Text #4

Max Range: 0.4598. Min Range: -0.4598

Max Act: 0.3170. Min Act: -0.0000

Data Index: 370729 (C4 (Web Text))

Max Activating Token Index: 356

Text #5

Max Range: 0.4598. Min Range: -0.4598

Max Act: 0.2785. Min Act: -0.0000

Data Index: 487256 (C4 (Web Text))

Max Activating Token Index: 302

Text #6

Max Range: 0.4598. Min Range: -0.4598

Max Act: 0.3136. Min Act: -0.0000

Data Index: 1312257 (C4 (Web Text))

Max Activating Token Index: 701

Text #7

Max Range: 0.4598. Min Range: -0.4598

Max Act: 0.2902. Min Act: -0.0001

Data Index: 92570 (C4 (Web Text))

Max Activating Token Index: 232

Text #8

Max Range: 0.4598. Min Range: -0.4598

Max Act: 0.2465. Min Act: -0.0001

Data Index: 294452 (C4 (Web Text))

Max Activating Token Index: 987

Text #9

Max Range: 0.4598. Min Range: -0.4598

Max Act: 0.2405. Min Act: -0.0000

Data Index: 817074 (C4 (Web Text))

Max Activating Token Index: 795

Text #10

Max Range: 0.4598. Min Range: -0.4598

Max Act: 0.2725. Min Act: -0.0000

Data Index: 6386 (C4 (Web Text))

Max Activating Token Index: 164

Text #11

Max Range: 0.4598. Min Range: -0.4598

Max Act: 0.2779. Min Act: -0.0000

Data Index: 944495 (C4 (Web Text))

Max Activating Token Index: 127

Text #12

Max Range: 0.4598. Min Range: -0.4598

Max Act: 0.2679. Min Act: -0.0000

Data Index: 999544 (C4 (Web Text))

Max Activating Token Index: 382

Text #13

Max Range: 0.4598. Min Range: -0.4598

Max Act: 0.2121. Min Act: -0.0001

Data Index: 1353899 (C4 (Web Text))

Max Activating Token Index: 820

Text #14

Max Range: 0.4598. Min Range: -0.4598

Max Act: 0.2475. Min Act: -0.0001

Data Index: 1553103 (Python Code)

Max Activating Token Index: 846

Text #15

Max Range: 0.4598. Min Range: -0.4598

Max Act: 0.2136. Min Act: -0.0001

Data Index: 974107 (C4 (Web Text))

Max Activating Token Index: 107

Text #16

Max Range: 0.4598. Min Range: -0.4598

Max Act: 0.2568. Min Act: -0.0001

Data Index: 569258 (C4 (Web Text))

Max Activating Token Index: 934

Text #17

Max Range: 0.4598. Min Range: -0.4598

Max Act: 0.2392. Min Act: -0.0001

Data Index: 283147 (C4 (Web Text))

Max Activating Token Index: 500

Text #18

Max Range: 0.4598. Min Range: -0.4598

Max Act: 0.1907. Min Act: -0.0000

Data Index: 1252090 (C4 (Web Text))

Max Activating Token Index: 694

Text #19

Max Range: 0.4598. Min Range: -0.4598

Max Act: 0.2735. Min Act: -0.0001

Data Index: 657576 (C4 (Web Text))

Max Activating Token Index: 338