Model: SoLU Model: 12 Layers, 6144 Neurons per Layer

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

Neuron 3624 in Layer 8

Load this data into an Interactive Neuroscope

See Documentation here

Transformer Lens Loading: HookedTransformer.from_pretrained('solu-12l')

Text #0

Max Range: 3.5986. Min Range: -3.5986

Max Act: 3.5986. Min Act: -0.0000

Data Index: 810399 (C4 (Web Text))

Max Activating Token Index: 380

Text #1

Max Range: 3.5986. Min Range: -3.5986

Max Act: 3.5458. Min Act: -0.0000

Data Index: 1443060 (Python Code)

Max Activating Token Index: 561

Text #2

Max Range: 3.5986. Min Range: -3.5986

Max Act: 3.2831. Min Act: -0.0000

Data Index: 648911 (C4 (Web Text))

Max Activating Token Index: 342

Text #3

Max Range: 3.5986. Min Range: -3.5986

Max Act: 3.3070. Min Act: -0.0000

Data Index: 305397 (C4 (Web Text))

Max Activating Token Index: 237

Text #4

Max Range: 3.5986. Min Range: -3.5986

Max Act: 2.9431. Min Act: -0.0000

Data Index: 1522425 (Python Code)

Max Activating Token Index: 566

Text #5

Max Range: 3.5986. Min Range: -3.5986

Max Act: 3.1464. Min Act: -0.0000

Data Index: 288823 (C4 (Web Text))

Max Activating Token Index: 397

Text #6

Max Range: 3.5986. Min Range: -3.5986

Max Act: 2.9354. Min Act: -0.0000

Data Index: 1632744 (Python Code)

Max Activating Token Index: 634

Text #7

Max Range: 3.5986. Min Range: -3.5986

Max Act: 2.9600. Min Act: -0.0000

Data Index: 1542305 (Python Code)

Max Activating Token Index: 125

Text #8

Max Range: 3.5986. Min Range: -3.5986

Max Act: 2.7879. Min Act: -0.0000

Data Index: 1159777 (C4 (Web Text))

Max Activating Token Index: 875

Text #9

Max Range: 3.5986. Min Range: -3.5986

Max Act: 2.9223. Min Act: -0.0000

Data Index: 1259473 (C4 (Web Text))

Max Activating Token Index: 222

Text #10

Max Range: 3.5986. Min Range: -3.5986

Max Act: 2.7954. Min Act: -0.0000

Data Index: 1407475 (Python Code)

Max Activating Token Index: 874

Text #11

Max Range: 3.5986. Min Range: -3.5986

Max Act: 2.7746. Min Act: -0.0000

Data Index: 1581081 (Python Code)

Max Activating Token Index: 154

Text #12

Max Range: 3.5986. Min Range: -3.5986

Max Act: 2.9005. Min Act: -0.0000

Data Index: 1569730 (Python Code)

Max Activating Token Index: 370

Text #13

Max Range: 3.5986. Min Range: -3.5986

Max Act: 2.8559. Min Act: -0.0000

Data Index: 249746 (C4 (Web Text))

Max Activating Token Index: 315

Text #14

Max Range: 3.5986. Min Range: -3.5986

Max Act: 2.5479. Min Act: 0.0000

Data Index: 1585669 (Python Code)

Max Activating Token Index: 311

Text #15

Max Range: 3.5986. Min Range: -3.5986

Max Act: 2.8484. Min Act: -0.0000

Data Index: 1607779 (Python Code)

Max Activating Token Index: 43

Text #16

Max Range: 3.5986. Min Range: -3.5986

Max Act: 2.5316. Min Act: -0.0000

Data Index: 1406331 (Python Code)

Max Activating Token Index: 759

Text #17

Max Range: 3.5986. Min Range: -3.5986

Max Act: 2.8075. Min Act: -0.0000

Data Index: 1442320 (Python Code)

Max Activating Token Index: 186

Text #18

Max Range: 3.5986. Min Range: -3.5986

Max Act: 2.9242. Min Act: -0.0000

Data Index: 492603 (C4 (Web Text))

Max Activating Token Index: 688

Text #19

Max Range: 3.5986. Min Range: -3.5986

Max Act: 2.7788. Min Act: -0.0000

Data Index: 1622269 (Python Code)

Max Activating Token Index: 646