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

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

Neuron 4604 in Layer 9

Load this data into an Interactive Neuroscope

See Documentation here

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

Text #0

Max Range: 0.7614. Min Range: -0.7614

Max Act: 0.7614. Min Act: 0.0000

Data Index: 538503 (C4 (Web Text))

Max Activating Token Index: 119

Text #1

Max Range: 0.7614. Min Range: -0.7614

Max Act: 0.5813. Min Act: -0.0000

Data Index: 80124 (C4 (Web Text))

Max Activating Token Index: 165

Text #2

Max Range: 0.7614. Min Range: -0.7614

Max Act: 0.5098. Min Act: -0.0000

Data Index: 632045 (C4 (Web Text))

Max Activating Token Index: 900

Text #3

Max Range: 0.7614. Min Range: -0.7614

Max Act: 0.5174. Min Act: -0.0000

Data Index: 97974 (C4 (Web Text))

Max Activating Token Index: 196

Text #4

Max Range: 0.7614. Min Range: -0.7614

Max Act: 0.5153. Min Act: -0.0000

Data Index: 204947 (C4 (Web Text))

Max Activating Token Index: 295

Text #5

Max Range: 0.7614. Min Range: -0.7614

Max Act: 0.4714. Min Act: -0.0000

Data Index: 97409 (C4 (Web Text))

Max Activating Token Index: 70

Text #6

Max Range: 0.7614. Min Range: -0.7614

Max Act: 0.4818. Min Act: -0.0000

Data Index: 788571 (C4 (Web Text))

Max Activating Token Index: 746

Text #7

Max Range: 0.7614. Min Range: -0.7614

Max Act: 0.4543. Min Act: -0.0000

Data Index: 600479 (C4 (Web Text))

Max Activating Token Index: 765

Text #8

Max Range: 0.7614. Min Range: -0.7614

Max Act: 0.4031. Min Act: -0.0000

Data Index: 363042 (C4 (Web Text))

Max Activating Token Index: 683

Text #9

Max Range: 0.7614. Min Range: -0.7614

Max Act: 0.4443. Min Act: -0.0000

Data Index: 247947 (C4 (Web Text))

Max Activating Token Index: 666

Text #10

Max Range: 0.7614. Min Range: -0.7614

Max Act: 0.4439. Min Act: -0.0000

Data Index: 393651 (C4 (Web Text))

Max Activating Token Index: 300

Text #11

Max Range: 0.7614. Min Range: -0.7614

Max Act: 0.4009. Min Act: -0.0000

Data Index: 180226 (C4 (Web Text))

Max Activating Token Index: 197

Text #12

Max Range: 0.7614. Min Range: -0.7614

Max Act: 0.4720. Min Act: -0.0000

Data Index: 1071883 (C4 (Web Text))

Max Activating Token Index: 660

Text #13

Max Range: 0.7614. Min Range: -0.7614

Max Act: 0.4065. Min Act: -0.0000

Data Index: 477299 (C4 (Web Text))

Max Activating Token Index: 182

Text #14

Max Range: 0.7614. Min Range: -0.7614

Max Act: 0.4173. Min Act: -0.0000

Data Index: 209512 (C4 (Web Text))

Max Activating Token Index: 1000

Text #15

Max Range: 0.7614. Min Range: -0.7614

Max Act: 0.4186. Min Act: -0.0000

Data Index: 904288 (C4 (Web Text))

Max Activating Token Index: 693

Text #16

Max Range: 0.7614. Min Range: -0.7614

Max Act: 0.4215. Min Act: -0.0000

Data Index: 529928 (C4 (Web Text))

Max Activating Token Index: 540

Text #17

Max Range: 0.7614. Min Range: -0.7614

Max Act: 0.3976. Min Act: -0.0000

Data Index: 833292 (C4 (Web Text))

Max Activating Token Index: 942

Text #18

Max Range: 0.7614. Min Range: -0.7614

Max Act: 0.4319. Min Act: -0.0000

Data Index: 388458 (C4 (Web Text))

Max Activating Token Index: 272

Text #19

Max Range: 0.7614. Min Range: -0.7614

Max Act: 0.3979. Min Act: -0.0000

Data Index: 1064398 (C4 (Web Text))

Max Activating Token Index: 193