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

Dataset: The Pile

Neuron 301 in Layer 6

Load this data into an Interactive Neuroscope

See Documentation here

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

Text #0

Max Range: 4.8031. Min Range: -4.8031

Max Act: 4.7506. Min Act: -0.0000

Data Index: 519333 (The Pile)

Max Activating Token Index: 828

Text #1

Max Range: 4.8031. Min Range: -4.8031

Max Act: 4.3072. Min Act: -0.0000

Data Index: 322375 (The Pile)

Max Activating Token Index: 99

Text #2

Max Range: 4.8031. Min Range: -4.8031

Max Act: 4.8031. Min Act: -0.0000

Data Index: 768347 (The Pile)

Max Activating Token Index: 72

Text #3

Max Range: 4.8031. Min Range: -4.8031

Max Act: 3.4999. Min Act: -0.0000

Data Index: 1166434 (The Pile)

Max Activating Token Index: 126

Text #4

Max Range: 4.8031. Min Range: -4.8031

Max Act: 4.4121. Min Act: -0.0000

Data Index: 1938329 (The Pile)

Max Activating Token Index: 258

Text #5

Max Range: 4.8031. Min Range: -4.8031

Max Act: 4.1678. Min Act: -0.0000

Data Index: 93978 (The Pile)

Max Activating Token Index: 541

Text #6

Max Range: 4.8031. Min Range: -4.8031

Max Act: 3.9652. Min Act: -0.0000

Data Index: 798809 (The Pile)

Max Activating Token Index: 519

Text #7

Max Range: 4.8031. Min Range: -4.8031

Max Act: 4.0493. Min Act: -0.0000

Data Index: 387617 (The Pile)

Max Activating Token Index: 72

Text #8

Max Range: 4.8031. Min Range: -4.8031

Max Act: 3.9860. Min Act: -0.0000

Data Index: 1768693 (The Pile)

Max Activating Token Index: 301

Text #9

Max Range: 4.8031. Min Range: -4.8031

Max Act: 3.8151. Min Act: -0.0000

Data Index: 264114 (The Pile)

Max Activating Token Index: 399

Text #10

Max Range: 4.8031. Min Range: -4.8031

Max Act: 3.4227. Min Act: -0.0000

Data Index: 1531762 (The Pile)

Max Activating Token Index: 826

Text #11

Max Range: 4.8031. Min Range: -4.8031

Max Act: 3.7674. Min Act: -0.0000

Data Index: 1011796 (The Pile)

Max Activating Token Index: 363

Text #12

Max Range: 4.8031. Min Range: -4.8031

Max Act: 3.7449. Min Act: -0.0000

Data Index: 474606 (The Pile)

Max Activating Token Index: 363

Text #13

Max Range: 4.8031. Min Range: -4.8031

Max Act: 3.6449. Min Act: -0.0000

Data Index: 199434 (The Pile)

Max Activating Token Index: 952

Text #14

Max Range: 4.8031. Min Range: -4.8031

Max Act: 3.7838. Min Act: -0.0000

Data Index: 419479 (The Pile)

Max Activating Token Index: 209

Text #15

Max Range: 4.8031. Min Range: -4.8031

Max Act: 3.0466. Min Act: -0.0000

Data Index: 400338 (The Pile)

Max Activating Token Index: 964

Text #16

Max Range: 4.8031. Min Range: -4.8031

Max Act: 3.6075. Min Act: -0.0000

Data Index: 1339201 (The Pile)

Max Activating Token Index: 538

Text #17

Max Range: 4.8031. Min Range: -4.8031

Max Act: 3.3499. Min Act: -0.0000

Data Index: 1315605 (The Pile)

Max Activating Token Index: 894

Text #18

Max Range: 4.8031. Min Range: -4.8031

Max Act: 3.6736. Min Act: -0.0000

Data Index: 1418220 (The Pile)

Max Activating Token Index: 346

Text #19

Max Range: 4.8031. Min Range: -4.8031

Max Act: 3.5598. Min Act: -0.0000

Data Index: 1978285 (The Pile)

Max Activating Token Index: 808