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

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

Neuron 1976 in Layer 3

Load this data into an Interactive Neuroscope

See Documentation here

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

Text #0

Max Range: 0.5875. Min Range: -0.5875

Max Act: 0.5860. Min Act: -0.0001

Data Index: 1600210 (Python Code)

Max Activating Token Index: 336

Text #1

Max Range: 0.5875. Min Range: -0.5875

Max Act: 0.5444. Min Act: -0.0001

Data Index: 1598089 (Python Code)

Max Activating Token Index: 648

Text #2

Max Range: 0.5875. Min Range: -0.5875

Max Act: 0.5875. Min Act: -0.0001

Data Index: 1598423 (Python Code)

Max Activating Token Index: 699

Text #3

Max Range: 0.5875. Min Range: -0.5875

Max Act: 0.4952. Min Act: -0.0001

Data Index: 1403695 (Python Code)

Max Activating Token Index: 248

Text #4

Max Range: 0.5875. Min Range: -0.5875

Max Act: 0.5021. Min Act: -0.0001

Data Index: 251999 (C4 (Web Text))

Max Activating Token Index: 716

Text #5

Max Range: 0.5875. Min Range: -0.5875

Max Act: 0.4498. Min Act: -0.0001

Data Index: 1575225 (Python Code)

Max Activating Token Index: 413

Text #6

Max Range: 0.5875. Min Range: -0.5875

Max Act: 0.4263. Min Act: -0.0001

Data Index: 880049 (C4 (Web Text))

Max Activating Token Index: 327

Text #7

Max Range: 0.5875. Min Range: -0.5875

Max Act: 0.4306. Min Act: -0.0001

Data Index: 1111669 (C4 (Web Text))

Max Activating Token Index: 894

Text #8

Max Range: 0.5875. Min Range: -0.5875

Max Act: 0.4774. Min Act: -0.0001

Data Index: 1602836 (Python Code)

Max Activating Token Index: 854

Text #9

Max Range: 0.5875. Min Range: -0.5875

Max Act: 0.4162. Min Act: -0.0001

Data Index: 1356184 (C4 (Web Text))

Max Activating Token Index: 920

Text #10

Max Range: 0.5875. Min Range: -0.5875

Max Act: 0.4230. Min Act: -0.0001

Data Index: 519648 (C4 (Web Text))

Max Activating Token Index: 939

Text #11

Max Range: 0.5875. Min Range: -0.5875

Max Act: 0.4140. Min Act: -0.0001

Data Index: 1339542 (C4 (Web Text))

Max Activating Token Index: 828

Text #12

Max Range: 0.5875. Min Range: -0.5875

Max Act: 0.4377. Min Act: -0.0001

Data Index: 61113 (C4 (Web Text))

Max Activating Token Index: 250

Text #13

Max Range: 0.5875. Min Range: -0.5875

Max Act: 0.4092. Min Act: -0.0001

Data Index: 1291755 (C4 (Web Text))

Max Activating Token Index: 171

Text #14

Max Range: 0.5875. Min Range: -0.5875

Max Act: 0.4232. Min Act: -0.0001

Data Index: 1055062 (C4 (Web Text))

Max Activating Token Index: 941

Text #15

Max Range: 0.5875. Min Range: -0.5875

Max Act: 0.4237. Min Act: -0.0001

Data Index: 1523498 (Python Code)

Max Activating Token Index: 117

Text #16

Max Range: 0.5875. Min Range: -0.5875

Max Act: 0.3756. Min Act: -0.0001

Data Index: 714522 (C4 (Web Text))

Max Activating Token Index: 638

Text #17

Max Range: 0.5875. Min Range: -0.5875

Max Act: 0.3739. Min Act: -0.0001

Data Index: 1540283 (Python Code)

Max Activating Token Index: 96

Text #18

Max Range: 0.5875. Min Range: -0.5875

Max Act: 0.3847. Min Act: -0.0001

Data Index: 1349123 (C4 (Web Text))

Max Activating Token Index: 232

Text #19

Max Range: 0.5875. Min Range: -0.5875

Max Act: 0.3839. Min Act: -0.0001

Data Index: 225397 (C4 (Web Text))

Max Activating Token Index: 335