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

Dataset: The Pile

Neuron 3928 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: 7.1369. Min Range: -7.1369

Max Act: 7.1369. Min Act: -0.0000

Data Index: 821625 (The Pile)

Max Activating Token Index: 583

Text #1

Max Range: 7.1369. Min Range: -7.1369

Max Act: 6.9515. Min Act: -0.0000

Data Index: 705035 (The Pile)

Max Activating Token Index: 608

Text #2

Max Range: 7.1369. Min Range: -7.1369

Max Act: 6.8772. Min Act: -0.0000

Data Index: 1612265 (The Pile)

Max Activating Token Index: 772

Text #3

Max Range: 7.1369. Min Range: -7.1369

Max Act: 6.5074. Min Act: -0.0000

Data Index: 94612 (The Pile)

Max Activating Token Index: 721

Text #4

Max Range: 7.1369. Min Range: -7.1369

Max Act: 6.2375. Min Act: -0.0000

Data Index: 1911882 (The Pile)

Max Activating Token Index: 552

Text #5

Max Range: 7.1369. Min Range: -7.1369

Max Act: 6.6740. Min Act: -0.0000

Data Index: 643167 (The Pile)

Max Activating Token Index: 634

Text #6

Max Range: 7.1369. Min Range: -7.1369

Max Act: 6.7660. Min Act: -0.0000

Data Index: 204913 (The Pile)

Max Activating Token Index: 254

Text #7

Max Range: 7.1369. Min Range: -7.1369

Max Act: 5.9226. Min Act: -0.0000

Data Index: 1864353 (The Pile)

Max Activating Token Index: 917

Text #8

Max Range: 7.1369. Min Range: -7.1369

Max Act: 6.5682. Min Act: -0.0000

Data Index: 1994616 (The Pile)

Max Activating Token Index: 836

Text #9

Max Range: 7.1369. Min Range: -7.1369

Max Act: 6.3670. Min Act: -0.0000

Data Index: 537049 (The Pile)

Max Activating Token Index: 255

Text #10

Max Range: 7.1369. Min Range: -7.1369

Max Act: 6.1745. Min Act: -0.0000

Data Index: 823926 (The Pile)

Max Activating Token Index: 815

Text #11

Max Range: 7.1369. Min Range: -7.1369

Max Act: 5.8978. Min Act: -0.0000

Data Index: 972717 (The Pile)

Max Activating Token Index: 921

Text #12

Max Range: 7.1369. Min Range: -7.1369

Max Act: 5.5700. Min Act: -0.0000

Data Index: 1499923 (The Pile)

Max Activating Token Index: 503

Text #13

Max Range: 7.1369. Min Range: -7.1369

Max Act: 5.4542. Min Act: -0.0000

Data Index: 1352072 (The Pile)

Max Activating Token Index: 327

Text #14

Max Range: 7.1369. Min Range: -7.1369

Max Act: 5.6265. Min Act: -0.0000

Data Index: 190914 (The Pile)

Max Activating Token Index: 201

Text #15

Max Range: 7.1369. Min Range: -7.1369

Max Act: 5.3093. Min Act: -0.0000

Data Index: 1328817 (The Pile)

Max Activating Token Index: 990

Text #16

Max Range: 7.1369. Min Range: -7.1369

Max Act: 5.3037. Min Act: -0.0000

Data Index: 788633 (The Pile)

Max Activating Token Index: 246

Text #17

Max Range: 7.1369. Min Range: -7.1369

Max Act: 5.0432. Min Act: -0.0000

Data Index: 1503888 (The Pile)

Max Activating Token Index: 359

Text #18

Max Range: 7.1369. Min Range: -7.1369

Max Act: 5.0085. Min Act: -0.0000

Data Index: 270229 (The Pile)

Max Activating Token Index: 288

Text #19

Max Range: 7.1369. Min Range: -7.1369

Max Act: 5.0830. Min Act: -0.0000

Data Index: 1735081 (The Pile)

Max Activating Token Index: 390