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

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

Neuron 8 in Layer 3

Load this data into an Interactive Neuroscope

See Documentation here

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

Text #0

Max Range: 0.1249. Min Range: -0.1249

Max Act: 0.1137. Min Act: -0.0001

Data Index: 272822 (C4 (Web Text))

Max Activating Token Index: 152

Text #1

Max Range: 0.1249. Min Range: -0.1249

Max Act: 0.1151. Min Act: -0.0001

Data Index: 638793 (C4 (Web Text))

Max Activating Token Index: 286

Text #2

Max Range: 0.1249. Min Range: -0.1249

Max Act: 0.1249. Min Act: -0.0001

Data Index: 217649 (C4 (Web Text))

Max Activating Token Index: 304

Text #3

Max Range: 0.1249. Min Range: -0.1249

Max Act: 0.1095. Min Act: -0.0001

Data Index: 1628847 (Python Code)

Max Activating Token Index: 807

Text #4

Max Range: 0.1249. Min Range: -0.1249

Max Act: 0.1233. Min Act: -0.0001

Data Index: 1594532 (Python Code)

Max Activating Token Index: 224

Text #5

Max Range: 0.1249. Min Range: -0.1249

Max Act: 0.1000. Min Act: -0.0001

Data Index: 942987 (C4 (Web Text))

Max Activating Token Index: 668

Text #6

Max Range: 0.1249. Min Range: -0.1249

Max Act: 0.1010. Min Act: -0.0001

Data Index: 921838 (C4 (Web Text))

Max Activating Token Index: 56

Text #7

Max Range: 0.1249. Min Range: -0.1249

Max Act: 0.1054. Min Act: -0.0001

Data Index: 992766 (C4 (Web Text))

Max Activating Token Index: 679

Text #8

Max Range: 0.1249. Min Range: -0.1249

Max Act: 0.1050. Min Act: -0.0001

Data Index: 1006375 (C4 (Web Text))

Max Activating Token Index: 80

Text #9

Max Range: 0.1249. Min Range: -0.1249

Max Act: 0.1004. Min Act: -0.0001

Data Index: 326288 (C4 (Web Text))

Max Activating Token Index: 741

Text #10

Max Range: 0.1249. Min Range: -0.1249

Max Act: 0.0958. Min Act: -0.0001

Data Index: 235791 (C4 (Web Text))

Max Activating Token Index: 936

Text #11

Max Range: 0.1249. Min Range: -0.1249

Max Act: 0.0922. Min Act: -0.0001

Data Index: 296044 (C4 (Web Text))

Max Activating Token Index: 767

Text #12

Max Range: 0.1249. Min Range: -0.1249

Max Act: 0.0929. Min Act: -0.0001

Data Index: 1205406 (C4 (Web Text))

Max Activating Token Index: 739

Text #13

Max Range: 0.1249. Min Range: -0.1249

Max Act: 0.0998. Min Act: -0.0001

Data Index: 1109001 (C4 (Web Text))

Max Activating Token Index: 954

Text #14

Max Range: 0.1249. Min Range: -0.1249

Max Act: 0.0994. Min Act: -0.0001

Data Index: 865966 (C4 (Web Text))

Max Activating Token Index: 342

Text #15

Max Range: 0.1249. Min Range: -0.1249

Max Act: 0.0916. Min Act: -0.0001

Data Index: 1597296 (Python Code)

Max Activating Token Index: 400

Text #16

Max Range: 0.1249. Min Range: -0.1249

Max Act: 0.0973. Min Act: -0.0001

Data Index: 1183205 (C4 (Web Text))

Max Activating Token Index: 319

Text #17

Max Range: 0.1249. Min Range: -0.1249

Max Act: 0.0955. Min Act: -0.0001

Data Index: 615379 (C4 (Web Text))

Max Activating Token Index: 554

Text #18

Max Range: 0.1249. Min Range: -0.1249

Max Act: 0.0890. Min Act: -0.0001

Data Index: 1394898 (Python Code)

Max Activating Token Index: 933

Text #19

Max Range: 0.1249. Min Range: -0.1249

Max Act: 0.0880. Min Act: -0.0001

Data Index: 157756 (C4 (Web Text))

Max Activating Token Index: 99