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

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

Neuron 3425 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: 0.3046. Min Range: -0.3046

Max Act: 0.3046. Min Act: -0.0000

Data Index: 886511 (C4 (Web Text))

Max Activating Token Index: 674

Text #1

Max Range: 0.3046. Min Range: -0.3046

Max Act: 0.2077. Min Act: -0.0000

Data Index: 954770 (C4 (Web Text))

Max Activating Token Index: 913

Text #2

Max Range: 0.3046. Min Range: -0.3046

Max Act: 0.1861. Min Act: -0.0000

Data Index: 225366 (C4 (Web Text))

Max Activating Token Index: 401

Text #3

Max Range: 0.3046. Min Range: -0.3046

Max Act: 0.1378. Min Act: -0.0000

Data Index: 148532 (C4 (Web Text))

Max Activating Token Index: 143

Text #4

Max Range: 0.3046. Min Range: -0.3046

Max Act: 0.1325. Min Act: -0.0000

Data Index: 400334 (C4 (Web Text))

Max Activating Token Index: 510

Text #5

Max Range: 0.3046. Min Range: -0.3046

Max Act: 0.1327. Min Act: -0.0000

Data Index: 1267238 (C4 (Web Text))

Max Activating Token Index: 747

Text #6

Max Range: 0.3046. Min Range: -0.3046

Max Act: 0.1318. Min Act: -0.0000

Data Index: 1106628 (C4 (Web Text))

Max Activating Token Index: 455

Text #7

Max Range: 0.3046. Min Range: -0.3046

Max Act: 0.1318. Min Act: -0.0000

Data Index: 191400 (C4 (Web Text))

Max Activating Token Index: 628

Text #8

Max Range: 0.3046. Min Range: -0.3046

Max Act: 0.0926. Min Act: -0.0000

Data Index: 794147 (C4 (Web Text))

Max Activating Token Index: 233

Text #9

Max Range: 0.3046. Min Range: -0.3046

Max Act: 0.1155. Min Act: -0.0000

Data Index: 811468 (C4 (Web Text))

Max Activating Token Index: 592

Text #10

Max Range: 0.3046. Min Range: -0.3046

Max Act: 0.1148. Min Act: -0.0000

Data Index: 338689 (C4 (Web Text))

Max Activating Token Index: 908

Text #11

Max Range: 0.3046. Min Range: -0.3046

Max Act: 0.1133. Min Act: -0.0000

Data Index: 910350 (C4 (Web Text))

Max Activating Token Index: 80

Text #12

Max Range: 0.3046. Min Range: -0.3046

Max Act: 0.1132. Min Act: -0.0000

Data Index: 1317154 (C4 (Web Text))

Max Activating Token Index: 329

Text #13

Max Range: 0.3046. Min Range: -0.3046

Max Act: 0.0922. Min Act: -0.0000

Data Index: 854790 (C4 (Web Text))

Max Activating Token Index: 529

Text #14

Max Range: 0.3046. Min Range: -0.3046

Max Act: 0.1132. Min Act: -0.0000

Data Index: 1637382 (Python Code)

Max Activating Token Index: 625

Text #15

Max Range: 0.3046. Min Range: -0.3046

Max Act: 0.0977. Min Act: -0.0000

Data Index: 1227881 (C4 (Web Text))

Max Activating Token Index: 28

Text #16

Max Range: 0.3046. Min Range: -0.3046

Max Act: 0.1160. Min Act: -0.0000

Data Index: 1156873 (C4 (Web Text))

Max Activating Token Index: 393

Text #17

Max Range: 0.3046. Min Range: -0.3046

Max Act: 0.1001. Min Act: -0.0000

Data Index: 994629 (C4 (Web Text))

Max Activating Token Index: 728

Text #18

Max Range: 0.3046. Min Range: -0.3046

Max Act: 0.1059. Min Act: -0.0000

Data Index: 84212 (C4 (Web Text))

Max Activating Token Index: 334

Text #19

Max Range: 0.3046. Min Range: -0.3046

Max Act: 0.0814. Min Act: -0.0000

Data Index: 225904 (C4 (Web Text))

Max Activating Token Index: 247