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

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

Neuron 3303 in Layer 4

Load this data into an Interactive Neuroscope

See Documentation here

Transformer Lens Loading: HookedTransformer.from_pretrained('solu-10l')

Text #0

Max Range: 1.4061. Min Range: -1.4061

Max Act: 1.4061. Min Act: -0.0001

Data Index: 1278893 (C4 (Web Text))

Max Activating Token Index: 668

Text #1

Max Range: 1.4061. Min Range: -1.4061

Max Act: 1.0472. Min Act: -0.0001

Data Index: 268121 (C4 (Web Text))

Max Activating Token Index: 129

Text #2

Max Range: 1.4061. Min Range: -1.4061

Max Act: 1.0390. Min Act: -0.0000

Data Index: 657850 (C4 (Web Text))

Max Activating Token Index: 23

Text #3

Max Range: 1.4061. Min Range: -1.4061

Max Act: 0.9802. Min Act: -0.0001

Data Index: 1554056 (Python Code)

Max Activating Token Index: 605

Text #4

Max Range: 1.4061. Min Range: -1.4061

Max Act: 1.0456. Min Act: -0.0001

Data Index: 823960 (C4 (Web Text))

Max Activating Token Index: 418

Text #5

Max Range: 1.4061. Min Range: -1.4061

Max Act: 1.0032. Min Act: -0.0001

Data Index: 839001 (C4 (Web Text))

Max Activating Token Index: 56

Text #6

Max Range: 1.4061. Min Range: -1.4061

Max Act: 1.0273. Min Act: -0.0001

Data Index: 965459 (C4 (Web Text))

Max Activating Token Index: 776

Text #7

Max Range: 1.4061. Min Range: -1.4061

Max Act: 0.9592. Min Act: -0.0001

Data Index: 1453071 (Python Code)

Max Activating Token Index: 52

Text #8

Max Range: 1.4061. Min Range: -1.4061

Max Act: 0.9528. Min Act: -0.0001

Data Index: 440153 (C4 (Web Text))

Max Activating Token Index: 81

Text #9

Max Range: 1.4061. Min Range: -1.4061

Max Act: 0.9935. Min Act: -0.0001

Data Index: 1556429 (Python Code)

Max Activating Token Index: 721

Text #10

Max Range: 1.4061. Min Range: -1.4061

Max Act: 0.9880. Min Act: -0.0000

Data Index: 949184 (C4 (Web Text))

Max Activating Token Index: 746

Text #11

Max Range: 1.4061. Min Range: -1.4061

Max Act: 0.9279. Min Act: -0.0001

Data Index: 1656195 (Python Code)

Max Activating Token Index: 664

Text #12

Max Range: 1.4061. Min Range: -1.4061

Max Act: 0.9231. Min Act: -0.0001

Data Index: 1259568 (C4 (Web Text))

Max Activating Token Index: 39

Text #13

Max Range: 1.4061. Min Range: -1.4061

Max Act: 0.9559. Min Act: -0.0001

Data Index: 724253 (C4 (Web Text))

Max Activating Token Index: 426

Text #14

Max Range: 1.4061. Min Range: -1.4061

Max Act: 0.9978. Min Act: -0.0001

Data Index: 1499854 (Python Code)

Max Activating Token Index: 806

Text #15

Max Range: 1.4061. Min Range: -1.4061

Max Act: 0.9674. Min Act: -0.0001

Data Index: 1444907 (Python Code)

Max Activating Token Index: 20

Text #16

Max Range: 1.4061. Min Range: -1.4061

Max Act: 0.9404. Min Act: -0.0001

Data Index: 1656962 (Python Code)

Max Activating Token Index: 443

Text #17

Max Range: 1.4061. Min Range: -1.4061

Max Act: 0.9650. Min Act: -0.0000

Data Index: 1622206 (Python Code)

Max Activating Token Index: 298

Text #18

Max Range: 1.4061. Min Range: -1.4061

Max Act: 0.9586. Min Act: -0.0001

Data Index: 1656726 (Python Code)

Max Activating Token Index: 239

Text #19

Max Range: 1.4061. Min Range: -1.4061

Max Act: 0.8769. Min Act: -0.0001

Data Index: 560952 (C4 (Web Text))

Max Activating Token Index: 65