Model: SoLU Model: 12 Layers, 6144 Neurons per Layer

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

Neuron 4932 in Layer 6

Load this data into an Interactive Neuroscope

See Documentation here

Transformer Lens Loading: HookedTransformer.from_pretrained('solu-12l')

Text #0

Max Range: 3.1964. Min Range: -3.1964

Max Act: 3.1964. Min Act: -0.0000

Data Index: 35088 (C4 (Web Text))

Max Activating Token Index: 741

Text #1

Max Range: 3.1964. Min Range: -3.1964

Max Act: 2.7189. Min Act: -0.0000

Data Index: 1558921 (Python Code)

Max Activating Token Index: 239

Text #2

Max Range: 3.1964. Min Range: -3.1964

Max Act: 2.7095. Min Act: -0.0000

Data Index: 1176555 (C4 (Web Text))

Max Activating Token Index: 220

Text #3

Max Range: 3.1964. Min Range: -3.1964

Max Act: 2.7113. Min Act: -0.0000

Data Index: 1645008 (Python Code)

Max Activating Token Index: 39

Text #4

Max Range: 3.1964. Min Range: -3.1964

Max Act: 2.6218. Min Act: -0.0000

Data Index: 464373 (C4 (Web Text))

Max Activating Token Index: 683

Text #5

Max Range: 3.1964. Min Range: -3.1964

Max Act: 2.4189. Min Act: -0.0000

Data Index: 1159038 (C4 (Web Text))

Max Activating Token Index: 711

Text #6

Max Range: 3.1964. Min Range: -3.1964

Max Act: 2.3963. Min Act: -0.0000

Data Index: 1149047 (C4 (Web Text))

Max Activating Token Index: 205

Text #7

Max Range: 3.1964. Min Range: -3.1964

Max Act: 2.3757. Min Act: -0.0000

Data Index: 1085881 (C4 (Web Text))

Max Activating Token Index: 532

Text #8

Max Range: 3.1964. Min Range: -3.1964

Max Act: 2.4736. Min Act: -0.0000

Data Index: 425240 (C4 (Web Text))

Max Activating Token Index: 878

Text #9

Max Range: 3.1964. Min Range: -3.1964

Max Act: 2.3362. Min Act: -0.0000

Data Index: 1322840 (C4 (Web Text))

Max Activating Token Index: 208

Text #10

Max Range: 3.1964. Min Range: -3.1964

Max Act: 2.4468. Min Act: -0.0000

Data Index: 359133 (C4 (Web Text))

Max Activating Token Index: 896

Text #11

Max Range: 3.1964. Min Range: -3.1964

Max Act: 2.4614. Min Act: -0.0000

Data Index: 1579637 (Python Code)

Max Activating Token Index: 63

Text #12

Max Range: 3.1964. Min Range: -3.1964

Max Act: 2.3292. Min Act: -0.0000

Data Index: 1379986 (Python Code)

Max Activating Token Index: 576

Text #13

Max Range: 3.1964. Min Range: -3.1964

Max Act: 2.2264. Min Act: -0.0000

Data Index: 413192 (C4 (Web Text))

Max Activating Token Index: 860

Text #14

Max Range: 3.1964. Min Range: -3.1964

Max Act: 2.3121. Min Act: -0.0000

Data Index: 1434256 (Python Code)

Max Activating Token Index: 890

Text #15

Max Range: 3.1964. Min Range: -3.1964

Max Act: 2.2947. Min Act: -0.0000

Data Index: 1404422 (Python Code)

Max Activating Token Index: 900

Text #16

Max Range: 3.1964. Min Range: -3.1964

Max Act: 2.3748. Min Act: -0.0000

Data Index: 1481944 (Python Code)

Max Activating Token Index: 131

Text #17

Max Range: 3.1964. Min Range: -3.1964

Max Act: 2.2758. Min Act: -0.0000

Data Index: 221606 (C4 (Web Text))

Max Activating Token Index: 392

Text #18

Max Range: 3.1964. Min Range: -3.1964

Max Act: 2.0171. Min Act: -0.0000

Data Index: 1219902 (C4 (Web Text))

Max Activating Token Index: 766

Text #19

Max Range: 3.1964. Min Range: -3.1964

Max Act: 2.2422. Min Act: -0.0000

Data Index: 1627658 (Python Code)

Max Activating Token Index: 281