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

Dataset: The Pile

Neuron 695 in Layer 6

Load this data into an Interactive Neuroscope

See Documentation here

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

Text #0

Max Range: 8.3667. Min Range: -8.3667

Max Act: 8.3667. Min Act: -0.0000

Data Index: 667473 (The Pile)

Max Activating Token Index: 263

Text #1

Max Range: 8.3667. Min Range: -8.3667

Max Act: 7.7434. Min Act: -0.0000

Data Index: 1543288 (The Pile)

Max Activating Token Index: 573

Text #2

Max Range: 8.3667. Min Range: -8.3667

Max Act: 7.6287. Min Act: -0.0000

Data Index: 708157 (The Pile)

Max Activating Token Index: 252

Text #3

Max Range: 8.3667. Min Range: -8.3667

Max Act: 7.7993. Min Act: -0.0000

Data Index: 1953322 (The Pile)

Max Activating Token Index: 432

Text #4

Max Range: 8.3667. Min Range: -8.3667

Max Act: 7.6440. Min Act: -0.0000

Data Index: 1943044 (The Pile)

Max Activating Token Index: 399

Text #5

Max Range: 8.3667. Min Range: -8.3667

Max Act: 7.6313. Min Act: -0.0000

Data Index: 463927 (The Pile)

Max Activating Token Index: 792

Text #6

Max Range: 8.3667. Min Range: -8.3667

Max Act: 7.6240. Min Act: -0.0000

Data Index: 1141139 (The Pile)

Max Activating Token Index: 633

Text #7

Max Range: 8.3667. Min Range: -8.3667

Max Act: 7.7219. Min Act: -0.0000

Data Index: 408880 (The Pile)

Max Activating Token Index: 200

Text #8

Max Range: 8.3667. Min Range: -8.3667

Max Act: 7.8968. Min Act: -0.0000

Data Index: 628998 (The Pile)

Max Activating Token Index: 565

Text #9

Max Range: 8.3667. Min Range: -8.3667

Max Act: 7.6703. Min Act: -0.0000

Data Index: 1521039 (The Pile)

Max Activating Token Index: 342

Text #10

Max Range: 8.3667. Min Range: -8.3667

Max Act: 7.4217. Min Act: -0.0000

Data Index: 1339832 (The Pile)

Max Activating Token Index: 155

Text #11

Max Range: 8.3667. Min Range: -8.3667

Max Act: 7.8094. Min Act: -0.0000

Data Index: 916863 (The Pile)

Max Activating Token Index: 940

Text #12

Max Range: 8.3667. Min Range: -8.3667

Max Act: 7.2476. Min Act: -0.0000

Data Index: 1409141 (The Pile)

Max Activating Token Index: 480

Text #13

Max Range: 8.3667. Min Range: -8.3667

Max Act: 7.7989. Min Act: -0.0000

Data Index: 1529713 (The Pile)

Max Activating Token Index: 584

Text #14

Max Range: 8.3667. Min Range: -8.3667

Max Act: 7.3898. Min Act: -0.0000

Data Index: 218440 (The Pile)

Max Activating Token Index: 398

Text #15

Max Range: 8.3667. Min Range: -8.3667

Max Act: 7.1822. Min Act: -0.0000

Data Index: 1993723 (The Pile)

Max Activating Token Index: 125

Text #16

Max Range: 8.3667. Min Range: -8.3667

Max Act: 7.5376. Min Act: -0.0000

Data Index: 813801 (The Pile)

Max Activating Token Index: 255

Text #17

Max Range: 8.3667. Min Range: -8.3667

Max Act: 7.5236. Min Act: -0.0000

Data Index: 102758 (The Pile)

Max Activating Token Index: 862

Text #18

Max Range: 8.3667. Min Range: -8.3667

Max Act: 7.5047. Min Act: -0.0000

Data Index: 51448 (The Pile)

Max Activating Token Index: 984

Text #19

Max Range: 8.3667. Min Range: -8.3667

Max Act: 7.6587. Min Act: -0.0000

Data Index: 1461061 (The Pile)

Max Activating Token Index: 340