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

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

Neuron 623 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: 0.1879. Min Range: -0.1879

Max Act: 0.1819. Min Act: -0.0001

Data Index: 938149 (C4 (Web Text))

Max Activating Token Index: 134

Text #1

Max Range: 0.1879. Min Range: -0.1879

Max Act: 0.1879. Min Act: -0.0001

Data Index: 195380 (C4 (Web Text))

Max Activating Token Index: 577

Text #2

Max Range: 0.1879. Min Range: -0.1879

Max Act: 0.1637. Min Act: -0.0001

Data Index: 869856 (C4 (Web Text))

Max Activating Token Index: 570

Text #3

Max Range: 0.1879. Min Range: -0.1879

Max Act: 0.1570. Min Act: -0.0001

Data Index: 538450 (C4 (Web Text))

Max Activating Token Index: 764

Text #4

Max Range: 0.1879. Min Range: -0.1879

Max Act: 0.1628. Min Act: -0.0001

Data Index: 625466 (C4 (Web Text))

Max Activating Token Index: 174

Text #5

Max Range: 0.1879. Min Range: -0.1879

Max Act: 0.1533. Min Act: -0.0001

Data Index: 909059 (C4 (Web Text))

Max Activating Token Index: 340

Text #6

Max Range: 0.1879. Min Range: -0.1879

Max Act: 0.1466. Min Act: -0.0001

Data Index: 426194 (C4 (Web Text))

Max Activating Token Index: 836

Text #7

Max Range: 0.1879. Min Range: -0.1879

Max Act: 0.1305. Min Act: -0.0001

Data Index: 10741 (C4 (Web Text))

Max Activating Token Index: 814

Text #8

Max Range: 0.1879. Min Range: -0.1879

Max Act: 0.1564. Min Act: -0.0001

Data Index: 354500 (C4 (Web Text))

Max Activating Token Index: 693

Text #9

Max Range: 0.1879. Min Range: -0.1879

Max Act: 0.1746. Min Act: -0.0001

Data Index: 208257 (C4 (Web Text))

Max Activating Token Index: 216

Text #10

Max Range: 0.1879. Min Range: -0.1879

Max Act: 0.1623. Min Act: -0.0001

Data Index: 1472337 (Python Code)

Max Activating Token Index: 338

Text #11

Max Range: 0.1879. Min Range: -0.1879

Max Act: 0.1619. Min Act: -0.0001

Data Index: 429966 (C4 (Web Text))

Max Activating Token Index: 171

Text #12

Max Range: 0.1879. Min Range: -0.1879

Max Act: 0.1663. Min Act: -0.0001

Data Index: 333601 (C4 (Web Text))

Max Activating Token Index: 785

Text #13

Max Range: 0.1879. Min Range: -0.1879

Max Act: 0.1600. Min Act: -0.0001

Data Index: 268075 (C4 (Web Text))

Max Activating Token Index: 99

Text #14

Max Range: 0.1879. Min Range: -0.1879

Max Act: 0.1524. Min Act: -0.0001

Data Index: 91052 (C4 (Web Text))

Max Activating Token Index: 296

Text #15

Max Range: 0.1879. Min Range: -0.1879

Max Act: 0.1583. Min Act: -0.0001

Data Index: 54890 (C4 (Web Text))

Max Activating Token Index: 525

Text #16

Max Range: 0.1879. Min Range: -0.1879

Max Act: 0.1448. Min Act: -0.0001

Data Index: 967579 (C4 (Web Text))

Max Activating Token Index: 672

Text #17

Max Range: 0.1879. Min Range: -0.1879

Max Act: 0.1624. Min Act: -0.0001

Data Index: 1416570 (Python Code)

Max Activating Token Index: 793

Text #18

Max Range: 0.1879. Min Range: -0.1879

Max Act: 0.1420. Min Act: -0.0001

Data Index: 167987 (C4 (Web Text))

Max Activating Token Index: 1019

Text #19

Max Range: 0.1879. Min Range: -0.1879

Max Act: 0.1607. Min Act: -0.0001

Data Index: 296839 (C4 (Web Text))

Max Activating Token Index: 319