Model: SoLU Model: 3 Layers, 2048 Neurons per Layer

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

Neuron 459 in Layer 2

Load this data into an Interactive Neuroscope

See Documentation here

Transformer Lens Loading: HookedTransformer.from_pretrained('solu-3l')

Text #0

Max Range: 0.1063. Min Range: -0.1063

Max Act: 0.1021. Min Act: -0.0003

Data Index: 882105 (C4 (Web Text))

Max Activating Token Index: 216

Text #1

Max Range: 0.1063. Min Range: -0.1063

Max Act: 0.1063. Min Act: -0.0003

Data Index: 237352 (C4 (Web Text))

Max Activating Token Index: 218

Text #2

Max Range: 0.1063. Min Range: -0.1063

Max Act: 0.0997. Min Act: -0.0003

Data Index: 809854 (C4 (Web Text))

Max Activating Token Index: 241

Text #3

Max Range: 0.1063. Min Range: -0.1063

Max Act: 0.0981. Min Act: -0.0003

Data Index: 969814 (C4 (Web Text))

Max Activating Token Index: 189

Text #4

Max Range: 0.1063. Min Range: -0.1063

Max Act: 0.0978. Min Act: -0.0003

Data Index: 515280 (C4 (Web Text))

Max Activating Token Index: 572

Text #5

Max Range: 0.1063. Min Range: -0.1063

Max Act: 0.0934. Min Act: -0.0004

Data Index: 841220 (C4 (Web Text))

Max Activating Token Index: 179

Text #6

Max Range: 0.1063. Min Range: -0.1063

Max Act: 0.0913. Min Act: -0.0003

Data Index: 385815 (C4 (Web Text))

Max Activating Token Index: 94

Text #7

Max Range: 0.1063. Min Range: -0.1063

Max Act: 0.0906. Min Act: -0.0003

Data Index: 147184 (C4 (Web Text))

Max Activating Token Index: 93

Text #8

Max Range: 0.1063. Min Range: -0.1063

Max Act: 0.0898. Min Act: -0.0003

Data Index: 251798 (C4 (Web Text))

Max Activating Token Index: 602

Text #9

Max Range: 0.1063. Min Range: -0.1063

Max Act: 0.0916. Min Act: -0.0003

Data Index: 247852 (C4 (Web Text))

Max Activating Token Index: 918

Text #10

Max Range: 0.1063. Min Range: -0.1063

Max Act: 0.0870. Min Act: -0.0003

Data Index: 1470991 (Python Code)

Max Activating Token Index: 955

Text #11

Max Range: 0.1063. Min Range: -0.1063

Max Act: 0.0907. Min Act: -0.0002

Data Index: 464517 (C4 (Web Text))

Max Activating Token Index: 1013

Text #12

Max Range: 0.1063. Min Range: -0.1063

Max Act: 0.0852. Min Act: -0.0003

Data Index: 502073 (C4 (Web Text))

Max Activating Token Index: 241

Text #13

Max Range: 0.1063. Min Range: -0.1063

Max Act: 0.0839. Min Act: -0.0003

Data Index: 869843 (C4 (Web Text))

Max Activating Token Index: 204

Text #14

Max Range: 0.1063. Min Range: -0.1063

Max Act: 0.0846. Min Act: -0.0004

Data Index: 1030016 (C4 (Web Text))

Max Activating Token Index: 929

Text #15

Max Range: 0.1063. Min Range: -0.1063

Max Act: 0.0845. Min Act: -0.0003

Data Index: 639663 (C4 (Web Text))

Max Activating Token Index: 283

Text #16

Max Range: 0.1063. Min Range: -0.1063

Max Act: 0.0848. Min Act: -0.0003

Data Index: 191304 (C4 (Web Text))

Max Activating Token Index: 964

Text #17

Max Range: 0.1063. Min Range: -0.1063

Max Act: 0.0803. Min Act: -0.0003

Data Index: 983712 (C4 (Web Text))

Max Activating Token Index: 452

Text #18

Max Range: 0.1063. Min Range: -0.1063

Max Act: 0.0831. Min Act: -0.0003

Data Index: 1067344 (C4 (Web Text))

Max Activating Token Index: 138

Text #19

Max Range: 0.1063. Min Range: -0.1063

Max Act: 0.0824. Min Act: -0.0003

Data Index: 258966 (C4 (Web Text))

Max Activating Token Index: 148