Model: SoLU Model: 3 Layers, 2048 Neurons per Layer

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

Neuron 1412 in Layer 2

Load this data into an Interactive Neuroscope

See Documentation here

Transformer Lens Loading: HookedTransformer.from_pretrained('solu-3l')

Text #0

Max Range: 0.0965. Min Range: -0.0965

Max Act: 0.0965. Min Act: -0.0003

Data Index: 1001623 (C4 (Web Text))

Max Activating Token Index: 931

Text #1

Max Range: 0.0965. Min Range: -0.0965

Max Act: 0.0777. Min Act: -0.0004

Data Index: 816064 (C4 (Web Text))

Max Activating Token Index: 311

Text #2

Max Range: 0.0965. Min Range: -0.0965

Max Act: 0.0744. Min Act: -0.0003

Data Index: 1228220 (C4 (Web Text))

Max Activating Token Index: 25

Text #3

Max Range: 0.0965. Min Range: -0.0965

Max Act: 0.0781. Min Act: -0.0003

Data Index: 276730 (C4 (Web Text))

Max Activating Token Index: 15

Text #4

Max Range: 0.0965. Min Range: -0.0965

Max Act: 0.0654. Min Act: -0.0003

Data Index: 1075468 (C4 (Web Text))

Max Activating Token Index: 273

Text #5

Max Range: 0.0965. Min Range: -0.0965

Max Act: 0.0656. Min Act: -0.0003

Data Index: 814740 (C4 (Web Text))

Max Activating Token Index: 403

Text #6

Max Range: 0.0965. Min Range: -0.0965

Max Act: 0.0586. Min Act: -0.0003

Data Index: 456119 (C4 (Web Text))

Max Activating Token Index: 161

Text #7

Max Range: 0.0965. Min Range: -0.0965

Max Act: 0.0576. Min Act: -0.0003

Data Index: 1201720 (C4 (Web Text))

Max Activating Token Index: 846

Text #8

Max Range: 0.0965. Min Range: -0.0965

Max Act: 0.0629. Min Act: -0.0004

Data Index: 1060779 (C4 (Web Text))

Max Activating Token Index: 310

Text #9

Max Range: 0.0965. Min Range: -0.0965

Max Act: 0.0607. Min Act: -0.0003

Data Index: 37756 (C4 (Web Text))

Max Activating Token Index: 63

Text #10

Max Range: 0.0965. Min Range: -0.0965

Max Act: 0.0603. Min Act: -0.0003

Data Index: 442618 (C4 (Web Text))

Max Activating Token Index: 551

Text #11

Max Range: 0.0965. Min Range: -0.0965

Max Act: 0.0542. Min Act: -0.0003

Data Index: 758613 (C4 (Web Text))

Max Activating Token Index: 477

Text #12

Max Range: 0.0965. Min Range: -0.0965

Max Act: 0.0559. Min Act: -0.0003

Data Index: 1103671 (C4 (Web Text))

Max Activating Token Index: 952

Text #13

Max Range: 0.0965. Min Range: -0.0965

Max Act: 0.0589. Min Act: -0.0003

Data Index: 415202 (C4 (Web Text))

Max Activating Token Index: 204

Text #14

Max Range: 0.0965. Min Range: -0.0965

Max Act: 0.0502. Min Act: -0.0003

Data Index: 142708 (C4 (Web Text))

Max Activating Token Index: 481

Text #15

Max Range: 0.0965. Min Range: -0.0965

Max Act: 0.0578. Min Act: -0.0003

Data Index: 812862 (C4 (Web Text))

Max Activating Token Index: 359

Text #16

Max Range: 0.0965. Min Range: -0.0965

Max Act: 0.0623. Min Act: -0.0003

Data Index: 508630 (C4 (Web Text))

Max Activating Token Index: 864

Text #17

Max Range: 0.0965. Min Range: -0.0965

Max Act: 0.0569. Min Act: -0.0003

Data Index: 1424737 (Python Code)

Max Activating Token Index: 743

Text #18

Max Range: 0.0965. Min Range: -0.0965

Max Act: 0.0583. Min Act: -0.0003

Data Index: 1068550 (C4 (Web Text))

Max Activating Token Index: 596

Text #19

Max Range: 0.0965. Min Range: -0.0965

Max Act: 0.0514. Min Act: -0.0003

Data Index: 207386 (C4 (Web Text))

Max Activating Token Index: 382