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

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

Neuron 2821 in Layer 5

Load this data into an Interactive Neuroscope

See Documentation here

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

Text #0

Max Range: 0.3969. Min Range: -0.3969

Max Act: 0.3969. Min Act: -0.0001

Data Index: 572810 (C4 (Web Text))

Max Activating Token Index: 931

Text #1

Max Range: 0.3969. Min Range: -0.3969

Max Act: 0.3911. Min Act: -0.0001

Data Index: 981148 (C4 (Web Text))

Max Activating Token Index: 604

Text #2

Max Range: 0.3969. Min Range: -0.3969

Max Act: 0.3678. Min Act: -0.0001

Data Index: 256705 (C4 (Web Text))

Max Activating Token Index: 322

Text #3

Max Range: 0.3969. Min Range: -0.3969

Max Act: 0.3616. Min Act: -0.0001

Data Index: 250404 (C4 (Web Text))

Max Activating Token Index: 680

Text #4

Max Range: 0.3969. Min Range: -0.3969

Max Act: 0.3827. Min Act: -0.0001

Data Index: 995644 (C4 (Web Text))

Max Activating Token Index: 233

Text #5

Max Range: 0.3969. Min Range: -0.3969

Max Act: 0.3557. Min Act: -0.0000

Data Index: 1348162 (C4 (Web Text))

Max Activating Token Index: 410

Text #6

Max Range: 0.3969. Min Range: -0.3969

Max Act: 0.3672. Min Act: -0.0000

Data Index: 366032 (C4 (Web Text))

Max Activating Token Index: 163

Text #7

Max Range: 0.3969. Min Range: -0.3969

Max Act: 0.3633. Min Act: -0.0000

Data Index: 776812 (C4 (Web Text))

Max Activating Token Index: 935

Text #8

Max Range: 0.3969. Min Range: -0.3969

Max Act: 0.3493. Min Act: -0.0000

Data Index: 1150867 (C4 (Web Text))

Max Activating Token Index: 212

Text #9

Max Range: 0.3969. Min Range: -0.3969

Max Act: 0.3606. Min Act: -0.0000

Data Index: 283926 (C4 (Web Text))

Max Activating Token Index: 343

Text #10

Max Range: 0.3969. Min Range: -0.3969

Max Act: 0.3106. Min Act: -0.0000

Data Index: 1634010 (Python Code)

Max Activating Token Index: 59

Text #11

Max Range: 0.3969. Min Range: -0.3969

Max Act: 0.3189. Min Act: -0.0000

Data Index: 1106444 (C4 (Web Text))

Max Activating Token Index: 739

Text #12

Max Range: 0.3969. Min Range: -0.3969

Max Act: 0.3124. Min Act: -0.0001

Data Index: 383873 (C4 (Web Text))

Max Activating Token Index: 870

Text #13

Max Range: 0.3969. Min Range: -0.3969

Max Act: 0.2875. Min Act: -0.0000

Data Index: 1558968 (Python Code)

Max Activating Token Index: 32

Text #14

Max Range: 0.3969. Min Range: -0.3969

Max Act: 0.2956. Min Act: -0.0000

Data Index: 38350 (C4 (Web Text))

Max Activating Token Index: 301

Text #15

Max Range: 0.3969. Min Range: -0.3969

Max Act: 0.2764. Min Act: -0.0000

Data Index: 1389622 (Python Code)

Max Activating Token Index: 94

Text #16

Max Range: 0.3969. Min Range: -0.3969

Max Act: 0.2822. Min Act: -0.0001

Data Index: 81110 (C4 (Web Text))

Max Activating Token Index: 528

Text #17

Max Range: 0.3969. Min Range: -0.3969

Max Act: 0.2826. Min Act: -0.0001

Data Index: 322575 (C4 (Web Text))

Max Activating Token Index: 776

Text #18

Max Range: 0.3969. Min Range: -0.3969

Max Act: 0.2778. Min Act: -0.0000

Data Index: 609055 (C4 (Web Text))

Max Activating Token Index: 592

Text #19

Max Range: 0.3969. Min Range: -0.3969

Max Act: 0.2650. Min Act: -0.0000

Data Index: 548672 (C4 (Web Text))

Max Activating Token Index: 779