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

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

Neuron 4256 in Layer 8

Load this data into an Interactive Neuroscope

See Documentation here

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

Text #0

Max Range: 1.7300. Min Range: -1.7300

Max Act: 0.0537. Min Act: -0.0000

Data Index: 754126 (C4 (Web Text))

Max Activating Token Index: 939

Text #1

Max Range: 1.7300. Min Range: -1.7300

Max Act: 1.3238. Min Act: -0.0000

Data Index: 63440 (C4 (Web Text))

Max Activating Token Index: 993

Text #2

Max Range: 1.7300. Min Range: -1.7300

Max Act: 1.4321. Min Act: -0.0000

Data Index: 473967 (C4 (Web Text))

Max Activating Token Index: 180

Text #3

Max Range: 1.7300. Min Range: -1.7300

Max Act: 1.7300. Min Act: -0.0000

Data Index: 919747 (C4 (Web Text))

Max Activating Token Index: 559

Text #4

Max Range: 1.7300. Min Range: -1.7300

Max Act: 1.4433. Min Act: -0.0000

Data Index: 1176462 (C4 (Web Text))

Max Activating Token Index: 82

Text #5

Max Range: 1.7300. Min Range: -1.7300

Max Act: 1.4375. Min Act: -0.0000

Data Index: 1037444 (C4 (Web Text))

Max Activating Token Index: 119

Text #6

Max Range: 1.7300. Min Range: -1.7300

Max Act: 1.1791. Min Act: -0.0000

Data Index: 790876 (C4 (Web Text))

Max Activating Token Index: 567

Text #7

Max Range: 1.7300. Min Range: -1.7300

Max Act: 1.2176. Min Act: -0.0000

Data Index: 984131 (C4 (Web Text))

Max Activating Token Index: 313

Text #8

Max Range: 1.7300. Min Range: -1.7300

Max Act: 1.2430. Min Act: -0.0000

Data Index: 466183 (C4 (Web Text))

Max Activating Token Index: 448

Text #9

Max Range: 1.7300. Min Range: -1.7300

Max Act: 1.2619. Min Act: -0.0000

Data Index: 408214 (C4 (Web Text))

Max Activating Token Index: 241

Text #10

Max Range: 1.7300. Min Range: -1.7300

Max Act: 1.2027. Min Act: -0.0000

Data Index: 2992 (C4 (Web Text))

Max Activating Token Index: 144

Text #11

Max Range: 1.7300. Min Range: -1.7300

Max Act: 1.3968. Min Act: -0.0000

Data Index: 239334 (C4 (Web Text))

Max Activating Token Index: 183

Text #12

Max Range: 1.7300. Min Range: -1.7300

Max Act: 0.8586. Min Act: -0.0000

Data Index: 866927 (C4 (Web Text))

Max Activating Token Index: 742

Text #13

Max Range: 1.7300. Min Range: -1.7300

Max Act: 1.1764. Min Act: -0.0000

Data Index: 311580 (C4 (Web Text))

Max Activating Token Index: 565

Text #14

Max Range: 1.7300. Min Range: -1.7300

Max Act: 1.1137. Min Act: -0.0000

Data Index: 213873 (C4 (Web Text))

Max Activating Token Index: 918

Text #15

Max Range: 1.7300. Min Range: -1.7300

Max Act: 0.9637. Min Act: -0.0000

Data Index: 236257 (C4 (Web Text))

Max Activating Token Index: 189

Text #16

Max Range: 1.7300. Min Range: -1.7300

Max Act: 0.7055. Min Act: -0.0000

Data Index: 1653350 (Python Code)

Max Activating Token Index: 1020

Text #17

Max Range: 1.7300. Min Range: -1.7300

Max Act: 0.7722. Min Act: -0.0000

Data Index: 1107317 (C4 (Web Text))

Max Activating Token Index: 846

Text #18

Max Range: 1.7300. Min Range: -1.7300

Max Act: 0.8510. Min Act: -0.0000

Data Index: 1443236 (Python Code)

Max Activating Token Index: 982

Text #19

Max Range: 1.7300. Min Range: -1.7300

Max Act: 0.8460. Min Act: -0.0000

Data Index: 645864 (C4 (Web Text))

Max Activating Token Index: 260