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

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

Neuron 1630 in Layer 5

Load this data into an Interactive Neuroscope

See Documentation here

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

Text #0

Max Range: 0.1550. Min Range: -0.1550

Max Act: 0.1550. Min Act: -0.0001

Data Index: 1617850 (Python Code)

Max Activating Token Index: 375

Text #1

Max Range: 0.1550. Min Range: -0.1550

Max Act: 0.1358. Min Act: 0.0001

Data Index: 25048 (C4 (Web Text))

Max Activating Token Index: 66

Text #2

Max Range: 0.1550. Min Range: -0.1550

Max Act: 0.1262. Min Act: -0.0000

Data Index: 851116 (C4 (Web Text))

Max Activating Token Index: 513

Text #3

Max Range: 0.1550. Min Range: -0.1550

Max Act: 0.1275. Min Act: 0.0000

Data Index: 167386 (C4 (Web Text))

Max Activating Token Index: 133

Text #4

Max Range: 0.1550. Min Range: -0.1550

Max Act: 0.1233. Min Act: -0.0000

Data Index: 62136 (C4 (Web Text))

Max Activating Token Index: 464

Text #5

Max Range: 0.1550. Min Range: -0.1550

Max Act: 0.1109. Min Act: 0.0001

Data Index: 71715 (C4 (Web Text))

Max Activating Token Index: 127

Text #6

Max Range: 0.1550. Min Range: -0.1550

Max Act: 0.1106. Min Act: 0.0000

Data Index: 125732 (C4 (Web Text))

Max Activating Token Index: 498

Text #7

Max Range: 0.1550. Min Range: -0.1550

Max Act: 0.1073. Min Act: 0.0001

Data Index: 275193 (C4 (Web Text))

Max Activating Token Index: 111

Text #8

Max Range: 0.1550. Min Range: -0.1550

Max Act: 0.1073. Min Act: -0.0001

Data Index: 520900 (C4 (Web Text))

Max Activating Token Index: 169

Text #9

Max Range: 0.1550. Min Range: -0.1550

Max Act: 0.1078. Min Act: 0.0001

Data Index: 315355 (C4 (Web Text))

Max Activating Token Index: 380

Text #10

Max Range: 0.1550. Min Range: -0.1550

Max Act: 0.1061. Min Act: -0.0001

Data Index: 257038 (C4 (Web Text))

Max Activating Token Index: 773

Text #11

Max Range: 0.1550. Min Range: -0.1550

Max Act: 0.1009. Min Act: 0.0000

Data Index: 367263 (C4 (Web Text))

Max Activating Token Index: 161

Text #12

Max Range: 0.1550. Min Range: -0.1550

Max Act: 0.0937. Min Act: -0.0000

Data Index: 38857 (C4 (Web Text))

Max Activating Token Index: 351

Text #13

Max Range: 0.1550. Min Range: -0.1550

Max Act: 0.1048. Min Act: -0.0000

Data Index: 262859 (C4 (Web Text))

Max Activating Token Index: 452

Text #14

Max Range: 0.1550. Min Range: -0.1550

Max Act: 0.1029. Min Act: -0.0000

Data Index: 98908 (C4 (Web Text))

Max Activating Token Index: 35

Text #15

Max Range: 0.1550. Min Range: -0.1550

Max Act: 0.0940. Min Act: 0.0000

Data Index: 50972 (C4 (Web Text))

Max Activating Token Index: 34

Text #16

Max Range: 0.1550. Min Range: -0.1550

Max Act: 0.1028. Min Act: 0.0000

Data Index: 733693 (C4 (Web Text))

Max Activating Token Index: 190

Text #17

Max Range: 0.1550. Min Range: -0.1550

Max Act: 0.0994. Min Act: -0.0000

Data Index: 425559 (C4 (Web Text))

Max Activating Token Index: 810

Text #18

Max Range: 0.1550. Min Range: -0.1550

Max Act: 0.0927. Min Act: -0.0000

Data Index: 1074024 (C4 (Web Text))

Max Activating Token Index: 510

Text #19

Max Range: 0.1550. Min Range: -0.1550

Max Act: 0.0951. Min Act: -0.0001

Data Index: 1370281 (Python Code)

Max Activating Token Index: 399