Model: SoLU Model: 10 Layers, 5120 Neurons per Layer

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

Neuron 4639 in Layer 7

Load this data into an Interactive Neuroscope

See Documentation here

Transformer Lens Loading: HookedTransformer.from_pretrained('solu-10l')

Text #0

Max Range: 2.0409. Min Range: -2.0409

Max Act: 2.0409. Min Act: -0.0000

Data Index: 863536 (C4 (Web Text))

Max Activating Token Index: 740

Text #1

Max Range: 2.0409. Min Range: -2.0409

Max Act: 1.6869. Min Act: -0.0000

Data Index: 1280152 (C4 (Web Text))

Max Activating Token Index: 580

Text #2

Max Range: 2.0409. Min Range: -2.0409

Max Act: 1.5851. Min Act: -0.0000

Data Index: 1076477 (C4 (Web Text))

Max Activating Token Index: 368

Text #3

Max Range: 2.0409. Min Range: -2.0409

Max Act: 1.5496. Min Act: -0.0000

Data Index: 965864 (C4 (Web Text))

Max Activating Token Index: 325

Text #4

Max Range: 2.0409. Min Range: -2.0409

Max Act: 1.5312. Min Act: -0.0000

Data Index: 1150570 (C4 (Web Text))

Max Activating Token Index: 772

Text #5

Max Range: 2.0409. Min Range: -2.0409

Max Act: 1.5242. Min Act: -0.0000

Data Index: 173400 (C4 (Web Text))

Max Activating Token Index: 259

Text #6

Max Range: 2.0409. Min Range: -2.0409

Max Act: 1.4923. Min Act: 0.0002

Data Index: 844689 (C4 (Web Text))

Max Activating Token Index: 211

Text #7

Max Range: 2.0409. Min Range: -2.0409

Max Act: 1.4860. Min Act: 0.0001

Data Index: 644684 (C4 (Web Text))

Max Activating Token Index: 215

Text #8

Max Range: 2.0409. Min Range: -2.0409

Max Act: 1.4722. Min Act: -0.0000

Data Index: 242012 (C4 (Web Text))

Max Activating Token Index: 393

Text #9

Max Range: 2.0409. Min Range: -2.0409

Max Act: 1.4405. Min Act: -0.0000

Data Index: 1298810 (C4 (Web Text))

Max Activating Token Index: 973

Text #10

Max Range: 2.0409. Min Range: -2.0409

Max Act: 1.5100. Min Act: -0.0000

Data Index: 385605 (C4 (Web Text))

Max Activating Token Index: 910

Text #11

Max Range: 2.0409. Min Range: -2.0409

Max Act: 1.3341. Min Act: 0.0001

Data Index: 1063988 (C4 (Web Text))

Max Activating Token Index: 296

Text #12

Max Range: 2.0409. Min Range: -2.0409

Max Act: 1.3213. Min Act: -0.0000

Data Index: 184619 (C4 (Web Text))

Max Activating Token Index: 414

Text #13

Max Range: 2.0409. Min Range: -2.0409

Max Act: 1.3440. Min Act: 0.0000

Data Index: 781931 (C4 (Web Text))

Max Activating Token Index: 936

Text #14

Max Range: 2.0409. Min Range: -2.0409

Max Act: 1.3450. Min Act: -0.0000

Data Index: 380681 (C4 (Web Text))

Max Activating Token Index: 154

Text #15

Max Range: 2.0409. Min Range: -2.0409

Max Act: 1.4295. Min Act: 0.0001

Data Index: 544842 (C4 (Web Text))

Max Activating Token Index: 33

Text #16

Max Range: 2.0409. Min Range: -2.0409

Max Act: 1.4174. Min Act: -0.0000

Data Index: 1596112 (Python Code)

Max Activating Token Index: 46

Text #17

Max Range: 2.0409. Min Range: -2.0409

Max Act: 1.3899. Min Act: -0.0000

Data Index: 177575 (C4 (Web Text))

Max Activating Token Index: 711

Text #18

Max Range: 2.0409. Min Range: -2.0409

Max Act: 1.3034. Min Act: -0.0000

Data Index: 755745 (C4 (Web Text))

Max Activating Token Index: 538

Text #19

Max Range: 2.0409. Min Range: -2.0409

Max Act: 1.2982. Min Act: -0.0000

Data Index: 449634 (C4 (Web Text))

Max Activating Token Index: 475