Model: SoLU Model: 4 Layers, 2048 Neurons per Layer

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

Neuron 437 in Layer 3

Load this data into an Interactive Neuroscope

See Documentation here

Transformer Lens Loading: HookedTransformer.from_pretrained('solu-4l')

Text #0

Max Range: 0.1222. Min Range: -0.1222

Max Act: 0.1187. Min Act: -0.0002

Data Index: 20161 (C4 (Web Text))

Max Activating Token Index: 326

Text #1

Max Range: 0.1222. Min Range: -0.1222

Max Act: 0.1185. Min Act: -0.0002

Data Index: 228774 (C4 (Web Text))

Max Activating Token Index: 878

Text #2

Max Range: 0.1222. Min Range: -0.1222

Max Act: 0.1222. Min Act: -0.0002

Data Index: 336847 (C4 (Web Text))

Max Activating Token Index: 721

Text #3

Max Range: 0.1222. Min Range: -0.1222

Max Act: 0.1166. Min Act: -0.0002

Data Index: 712578 (C4 (Web Text))

Max Activating Token Index: 430

Text #4

Max Range: 0.1222. Min Range: -0.1222

Max Act: 0.1143. Min Act: -0.0002

Data Index: 151901 (C4 (Web Text))

Max Activating Token Index: 296

Text #5

Max Range: 0.1222. Min Range: -0.1222

Max Act: 0.1054. Min Act: -0.0002

Data Index: 207546 (C4 (Web Text))

Max Activating Token Index: 475

Text #6

Max Range: 0.1222. Min Range: -0.1222

Max Act: 0.1081. Min Act: -0.0002

Data Index: 922398 (C4 (Web Text))

Max Activating Token Index: 866

Text #7

Max Range: 0.1222. Min Range: -0.1222

Max Act: 0.1027. Min Act: -0.0002

Data Index: 1214021 (C4 (Web Text))

Max Activating Token Index: 53

Text #8

Max Range: 0.1222. Min Range: -0.1222

Max Act: 0.1019. Min Act: -0.0002

Data Index: 692131 (C4 (Web Text))

Max Activating Token Index: 484

Text #9

Max Range: 0.1222. Min Range: -0.1222

Max Act: 0.0962. Min Act: -0.0002

Data Index: 876679 (C4 (Web Text))

Max Activating Token Index: 590

Text #10

Max Range: 0.1222. Min Range: -0.1222

Max Act: 0.0972. Min Act: -0.0001

Data Index: 92925 (C4 (Web Text))

Max Activating Token Index: 202

Text #11

Max Range: 0.1222. Min Range: -0.1222

Max Act: 0.0936. Min Act: -0.0002

Data Index: 1388596 (Python Code)

Max Activating Token Index: 888

Text #12

Max Range: 0.1222. Min Range: -0.1222

Max Act: 0.0976. Min Act: -0.0002

Data Index: 91565 (C4 (Web Text))

Max Activating Token Index: 516

Text #13

Max Range: 0.1222. Min Range: -0.1222

Max Act: 0.0975. Min Act: -0.0002

Data Index: 12326 (C4 (Web Text))

Max Activating Token Index: 39

Text #14

Max Range: 0.1222. Min Range: -0.1222

Max Act: 0.0965. Min Act: -0.0002

Data Index: 333424 (C4 (Web Text))

Max Activating Token Index: 69

Text #15

Max Range: 0.1222. Min Range: -0.1222

Max Act: 0.0929. Min Act: -0.0002

Data Index: 870306 (C4 (Web Text))

Max Activating Token Index: 114

Text #16

Max Range: 0.1222. Min Range: -0.1222

Max Act: 0.0925. Min Act: -0.0002

Data Index: 540906 (C4 (Web Text))

Max Activating Token Index: 1007

Text #17

Max Range: 0.1222. Min Range: -0.1222

Max Act: 0.0953. Min Act: -0.0002

Data Index: 515945 (C4 (Web Text))

Max Activating Token Index: 810

Text #18

Max Range: 0.1222. Min Range: -0.1222

Max Act: 0.0967. Min Act: -0.0002

Data Index: 140061 (C4 (Web Text))

Max Activating Token Index: 564

Text #19

Max Range: 0.1222. Min Range: -0.1222

Max Act: 0.0903. Min Act: -0.0002

Data Index: 1057378 (C4 (Web Text))

Max Activating Token Index: 526