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

Dataset: The Pile

Neuron 3895 in Layer 7

Load this data into an Interactive Neuroscope

See Documentation here

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

Text #0

Max Range: 5.6554. Min Range: -5.6554

Max Act: 5.5858. Min Act: -0.0000

Data Index: 408760 (The Pile)

Max Activating Token Index: 174

Text #1

Max Range: 5.6554. Min Range: -5.6554

Max Act: 5.6325. Min Act: -0.0000

Data Index: 1986126 (The Pile)

Max Activating Token Index: 292

Text #2

Max Range: 5.6554. Min Range: -5.6554

Max Act: 5.6554. Min Act: -0.0000

Data Index: 1403237 (The Pile)

Max Activating Token Index: 731

Text #3

Max Range: 5.6554. Min Range: -5.6554

Max Act: 5.4866. Min Act: -0.0000

Data Index: 1718905 (The Pile)

Max Activating Token Index: 134

Text #4

Max Range: 5.6554. Min Range: -5.6554

Max Act: 5.1560. Min Act: -0.0000

Data Index: 168435 (The Pile)

Max Activating Token Index: 92

Text #5

Max Range: 5.6554. Min Range: -5.6554

Max Act: 5.1960. Min Act: -0.0000

Data Index: 1726581 (The Pile)

Max Activating Token Index: 283

Text #6

Max Range: 5.6554. Min Range: -5.6554

Max Act: 5.2387. Min Act: -0.0000

Data Index: 29345 (The Pile)

Max Activating Token Index: 600

Text #7

Max Range: 5.6554. Min Range: -5.6554

Max Act: 5.2568. Min Act: -0.0000

Data Index: 1932527 (The Pile)

Max Activating Token Index: 313

Text #8

Max Range: 5.6554. Min Range: -5.6554

Max Act: 5.3509. Min Act: -0.0000

Data Index: 79963 (The Pile)

Max Activating Token Index: 775

Text #9

Max Range: 5.6554. Min Range: -5.6554

Max Act: 5.0471. Min Act: -0.0000

Data Index: 1236096 (The Pile)

Max Activating Token Index: 356

Text #10

Max Range: 5.6554. Min Range: -5.6554

Max Act: 5.1994. Min Act: -0.0000

Data Index: 401413 (The Pile)

Max Activating Token Index: 555

Text #11

Max Range: 5.6554. Min Range: -5.6554

Max Act: 5.0984. Min Act: -0.0000

Data Index: 347498 (The Pile)

Max Activating Token Index: 302

Text #12

Max Range: 5.6554. Min Range: -5.6554

Max Act: 4.9291. Min Act: -0.0000

Data Index: 704970 (The Pile)

Max Activating Token Index: 408

Text #13

Max Range: 5.6554. Min Range: -5.6554

Max Act: 5.0986. Min Act: -0.0001

Data Index: 287902 (The Pile)

Max Activating Token Index: 222

Text #14

Max Range: 5.6554. Min Range: -5.6554

Max Act: 5.3224. Min Act: -0.0000

Data Index: 454294 (The Pile)

Max Activating Token Index: 307

Text #15

Max Range: 5.6554. Min Range: -5.6554

Max Act: 5.2315. Min Act: -0.0000

Data Index: 1834764 (The Pile)

Max Activating Token Index: 276

Text #16

Max Range: 5.6554. Min Range: -5.6554

Max Act: 5.2375. Min Act: -0.0000

Data Index: 643819 (The Pile)

Max Activating Token Index: 893

Text #17

Max Range: 5.6554. Min Range: -5.6554

Max Act: 4.9133. Min Act: -0.0000

Data Index: 1484688 (The Pile)

Max Activating Token Index: 446

Text #18

Max Range: 5.6554. Min Range: -5.6554

Max Act: 4.5758. Min Act: -0.0000

Data Index: 203561 (The Pile)

Max Activating Token Index: 771

Text #19

Max Range: 5.6554. Min Range: -5.6554

Max Act: 4.8461. Min Act: -0.0000

Data Index: 1108961 (The Pile)

Max Activating Token Index: 649