Model: SoLU Model: 6 Layers, 3072 Neurons per Layer

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

Neuron 3049 in Layer 4

Load this data into an Interactive Neuroscope

See Documentation here

Transformer Lens Loading: HookedTransformer.from_pretrained('solu-6l')

Text #0

Max Range: 0.2539. Min Range: -0.2539

Max Act: 0.2539. Min Act: -0.0001

Data Index: 1485967 (Python Code)

Max Activating Token Index: 990

Text #1

Max Range: 0.2539. Min Range: -0.2539

Max Act: 0.2396. Min Act: -0.0001

Data Index: 478467 (C4 (Web Text))

Max Activating Token Index: 334

Text #2

Max Range: 0.2539. Min Range: -0.2539

Max Act: 0.2386. Min Act: -0.0001

Data Index: 37182 (C4 (Web Text))

Max Activating Token Index: 723

Text #3

Max Range: 0.2539. Min Range: -0.2539

Max Act: 0.2336. Min Act: -0.0001

Data Index: 837528 (C4 (Web Text))

Max Activating Token Index: 807

Text #4

Max Range: 0.2539. Min Range: -0.2539

Max Act: 0.2278. Min Act: -0.0001

Data Index: 26007 (C4 (Web Text))

Max Activating Token Index: 622

Text #5

Max Range: 0.2539. Min Range: -0.2539

Max Act: 0.2202. Min Act: -0.0001

Data Index: 427464 (C4 (Web Text))

Max Activating Token Index: 981

Text #6

Max Range: 0.2539. Min Range: -0.2539

Max Act: 0.2094. Min Act: -0.0001

Data Index: 924720 (C4 (Web Text))

Max Activating Token Index: 370

Text #7

Max Range: 0.2539. Min Range: -0.2539

Max Act: 0.2155. Min Act: -0.0001

Data Index: 1113744 (C4 (Web Text))

Max Activating Token Index: 642

Text #8

Max Range: 0.2539. Min Range: -0.2539

Max Act: 0.2257. Min Act: -0.0001

Data Index: 332675 (C4 (Web Text))

Max Activating Token Index: 212

Text #9

Max Range: 0.2539. Min Range: -0.2539

Max Act: 0.2095. Min Act: -0.0001

Data Index: 900230 (C4 (Web Text))

Max Activating Token Index: 359

Text #10

Max Range: 0.2539. Min Range: -0.2539

Max Act: 0.2162. Min Act: -0.0001

Data Index: 56191 (C4 (Web Text))

Max Activating Token Index: 620

Text #11

Max Range: 0.2539. Min Range: -0.2539

Max Act: 0.2158. Min Act: -0.0001

Data Index: 80373 (C4 (Web Text))

Max Activating Token Index: 862

Text #12

Max Range: 0.2539. Min Range: -0.2539

Max Act: 0.2085. Min Act: -0.0001

Data Index: 128317 (C4 (Web Text))

Max Activating Token Index: 358

Text #13

Max Range: 0.2539. Min Range: -0.2539

Max Act: 0.1975. Min Act: -0.0001

Data Index: 151244 (C4 (Web Text))

Max Activating Token Index: 297

Text #14

Max Range: 0.2539. Min Range: -0.2539

Max Act: 0.2045. Min Act: -0.0001

Data Index: 1096650 (C4 (Web Text))

Max Activating Token Index: 694

Text #15

Max Range: 0.2539. Min Range: -0.2539

Max Act: 0.2022. Min Act: -0.0001

Data Index: 745346 (C4 (Web Text))

Max Activating Token Index: 160

Text #16

Max Range: 0.2539. Min Range: -0.2539

Max Act: 0.1990. Min Act: -0.0001

Data Index: 298283 (C4 (Web Text))

Max Activating Token Index: 882

Text #17

Max Range: 0.2539. Min Range: -0.2539

Max Act: 0.1925. Min Act: -0.0001

Data Index: 1103004 (C4 (Web Text))

Max Activating Token Index: 869

Text #18

Max Range: 0.2539. Min Range: -0.2539

Max Act: 0.1923. Min Act: -0.0001

Data Index: 1355439 (C4 (Web Text))

Max Activating Token Index: 531

Text #19

Max Range: 0.2539. Min Range: -0.2539

Max Act: 0.1989. Min Act: -0.0001

Data Index: 990829 (C4 (Web Text))

Max Activating Token Index: 427