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

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

Neuron 4544 in Layer 6

Load this data into an Interactive Neuroscope

See Documentation here

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

Text #0

Max Range: 0.9616. Min Range: -0.9616

Max Act: 0.9616. Min Act: -0.0000

Data Index: 916829 (C4 (Web Text))

Max Activating Token Index: 26

Text #1

Max Range: 0.9616. Min Range: -0.9616

Max Act: 0.9173. Min Act: -0.0000

Data Index: 811584 (C4 (Web Text))

Max Activating Token Index: 53

Text #2

Max Range: 0.9616. Min Range: -0.9616

Max Act: 0.9022. Min Act: -0.0000

Data Index: 1021527 (C4 (Web Text))

Max Activating Token Index: 966

Text #3

Max Range: 0.9616. Min Range: -0.9616

Max Act: 0.8545. Min Act: -0.0000

Data Index: 389112 (C4 (Web Text))

Max Activating Token Index: 60

Text #4

Max Range: 0.9616. Min Range: -0.9616

Max Act: 0.8775. Min Act: -0.0000

Data Index: 1567564 (Python Code)

Max Activating Token Index: 32

Text #5

Max Range: 0.9616. Min Range: -0.9616

Max Act: 0.8161. Min Act: -0.0000

Data Index: 447385 (C4 (Web Text))

Max Activating Token Index: 224

Text #6

Max Range: 0.9616. Min Range: -0.9616

Max Act: 0.8478. Min Act: -0.0000

Data Index: 692578 (C4 (Web Text))

Max Activating Token Index: 133

Text #7

Max Range: 0.9616. Min Range: -0.9616

Max Act: 0.8449. Min Act: -0.0000

Data Index: 151173 (C4 (Web Text))

Max Activating Token Index: 21

Text #8

Max Range: 0.9616. Min Range: -0.9616

Max Act: 0.8184. Min Act: -0.0000

Data Index: 1042734 (C4 (Web Text))

Max Activating Token Index: 53

Text #9

Max Range: 0.9616. Min Range: -0.9616

Max Act: 0.8166. Min Act: -0.0000

Data Index: 785448 (C4 (Web Text))

Max Activating Token Index: 49

Text #10

Max Range: 0.9616. Min Range: -0.9616

Max Act: 0.8463. Min Act: -0.0000

Data Index: 1510220 (Python Code)

Max Activating Token Index: 76

Text #11

Max Range: 0.9616. Min Range: -0.9616

Max Act: 0.8372. Min Act: -0.0000

Data Index: 1524423 (Python Code)

Max Activating Token Index: 42

Text #12

Max Range: 0.9616. Min Range: -0.9616

Max Act: 0.8084. Min Act: -0.0000

Data Index: 730238 (C4 (Web Text))

Max Activating Token Index: 129

Text #13

Max Range: 0.9616. Min Range: -0.9616

Max Act: 0.8333. Min Act: -0.0000

Data Index: 55303 (C4 (Web Text))

Max Activating Token Index: 88

Text #14

Max Range: 0.9616. Min Range: -0.9616

Max Act: 0.8059. Min Act: -0.0000

Data Index: 1519334 (Python Code)

Max Activating Token Index: 205

Text #15

Max Range: 0.9616. Min Range: -0.9616

Max Act: 0.8286. Min Act: -0.0000

Data Index: 645963 (C4 (Web Text))

Max Activating Token Index: 100

Text #16

Max Range: 0.9616. Min Range: -0.9616

Max Act: 0.8260. Min Act: -0.0000

Data Index: 1359151 (C4 (Web Text))

Max Activating Token Index: 98

Text #17

Max Range: 0.9616. Min Range: -0.9616

Max Act: 0.8232. Min Act: -0.0000

Data Index: 574956 (C4 (Web Text))

Max Activating Token Index: 47

Text #18

Max Range: 0.9616. Min Range: -0.9616

Max Act: 0.8248. Min Act: -0.0000

Data Index: 506956 (C4 (Web Text))

Max Activating Token Index: 178

Text #19

Max Range: 0.9616. Min Range: -0.9616

Max Act: 0.8149. Min Act: -0.0000

Data Index: 10406 (C4 (Web Text))

Max Activating Token Index: 90