Model: SoLU Model: 8 Layers, 4096 Neurons per Layer

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

Neuron 4070 in Layer 6

Load this data into an Interactive Neuroscope

See Documentation here

Transformer Lens Loading: HookedTransformer.from_pretrained('solu-8l')

Text #0

Max Range: 0.0920. Min Range: -0.0920

Max Act: 0.0920. Min Act: -0.0001

Data Index: 978092 (C4 (Web Text))

Max Activating Token Index: 894

Text #1

Max Range: 0.0920. Min Range: -0.0920

Max Act: 0.0886. Min Act: -0.0001

Data Index: 924172 (C4 (Web Text))

Max Activating Token Index: 579

Text #2

Max Range: 0.0920. Min Range: -0.0920

Max Act: 0.0821. Min Act: -0.0000

Data Index: 728638 (C4 (Web Text))

Max Activating Token Index: 61

Text #3

Max Range: 0.0920. Min Range: -0.0920

Max Act: 0.0703. Min Act: -0.0000

Data Index: 461381 (C4 (Web Text))

Max Activating Token Index: 240

Text #4

Max Range: 0.0920. Min Range: -0.0920

Max Act: 0.0696. Min Act: -0.0001

Data Index: 1579243 (Python Code)

Max Activating Token Index: 384

Text #5

Max Range: 0.0920. Min Range: -0.0920

Max Act: 0.0686. Min Act: -0.0000

Data Index: 615280 (C4 (Web Text))

Max Activating Token Index: 602

Text #6

Max Range: 0.0920. Min Range: -0.0920

Max Act: 0.0677. Min Act: -0.0001

Data Index: 1405724 (Python Code)

Max Activating Token Index: 272

Text #7

Max Range: 0.0920. Min Range: -0.0920

Max Act: 0.0653. Min Act: -0.0000

Data Index: 174040 (C4 (Web Text))

Max Activating Token Index: 583

Text #8

Max Range: 0.0920. Min Range: -0.0920

Max Act: 0.0658. Min Act: -0.0000

Data Index: 793072 (C4 (Web Text))

Max Activating Token Index: 216

Text #9

Max Range: 0.0920. Min Range: -0.0920

Max Act: 0.0605. Min Act: -0.0000

Data Index: 330615 (C4 (Web Text))

Max Activating Token Index: 32

Text #10

Max Range: 0.0920. Min Range: -0.0920

Max Act: 0.0647. Min Act: -0.0000

Data Index: 1166013 (C4 (Web Text))

Max Activating Token Index: 724

Text #11

Max Range: 0.0920. Min Range: -0.0920

Max Act: 0.0580. Min Act: -0.0000

Data Index: 526514 (C4 (Web Text))

Max Activating Token Index: 286

Text #12

Max Range: 0.0920. Min Range: -0.0920

Max Act: 0.0630. Min Act: -0.0000

Data Index: 716293 (C4 (Web Text))

Max Activating Token Index: 547

Text #13

Max Range: 0.0920. Min Range: -0.0920

Max Act: 0.0566. Min Act: -0.0000

Data Index: 243335 (C4 (Web Text))

Max Activating Token Index: 352

Text #14

Max Range: 0.0920. Min Range: -0.0920

Max Act: 0.0539. Min Act: -0.0000

Data Index: 1277604 (C4 (Web Text))

Max Activating Token Index: 576

Text #15

Max Range: 0.0920. Min Range: -0.0920

Max Act: 0.0587. Min Act: -0.0000

Data Index: 1644688 (Python Code)

Max Activating Token Index: 221

Text #16

Max Range: 0.0920. Min Range: -0.0920

Max Act: 0.0601. Min Act: -0.0000

Data Index: 1320650 (C4 (Web Text))

Max Activating Token Index: 776

Text #17

Max Range: 0.0920. Min Range: -0.0920

Max Act: 0.0518. Min Act: -0.0000

Data Index: 961239 (C4 (Web Text))

Max Activating Token Index: 913

Text #18

Max Range: 0.0920. Min Range: -0.0920

Max Act: 0.0549. Min Act: -0.0000

Data Index: 239387 (C4 (Web Text))

Max Activating Token Index: 371

Text #19

Max Range: 0.0920. Min Range: -0.0920

Max Act: 0.0550. Min Act: -0.0001

Data Index: 1611503 (Python Code)

Max Activating Token Index: 166