Model: SoLU Model: 12 Layers, 6144 Neurons per Layer

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

Neuron 3515 in Layer 5

Load this data into an Interactive Neuroscope

See Documentation here

Transformer Lens Loading: HookedTransformer.from_pretrained('solu-12l')

Text #0

Max Range: 4.8155. Min Range: -4.8155

Max Act: 4.8155. Min Act: -0.0001

Data Index: 79730 (C4 (Web Text))

Max Activating Token Index: 753

Text #1

Max Range: 4.8155. Min Range: -4.8155

Max Act: 4.2931. Min Act: -0.0001

Data Index: 920062 (C4 (Web Text))

Max Activating Token Index: 442

Text #2

Max Range: 4.8155. Min Range: -4.8155

Max Act: 2.9443. Min Act: -0.0001

Data Index: 251913 (C4 (Web Text))

Max Activating Token Index: 206

Text #3

Max Range: 4.8155. Min Range: -4.8155

Max Act: 2.9597. Min Act: -0.0001

Data Index: 534215 (C4 (Web Text))

Max Activating Token Index: 918

Text #4

Max Range: 4.8155. Min Range: -4.8155

Max Act: 3.0326. Min Act: -0.0001

Data Index: 867380 (C4 (Web Text))

Max Activating Token Index: 824

Text #5

Max Range: 4.8155. Min Range: -4.8155

Max Act: 2.8895. Min Act: -0.0001

Data Index: 149181 (C4 (Web Text))

Max Activating Token Index: 806

Text #6

Max Range: 4.8155. Min Range: -4.8155

Max Act: 3.0723. Min Act: -0.0001

Data Index: 827486 (C4 (Web Text))

Max Activating Token Index: 958

Text #7

Max Range: 4.8155. Min Range: -4.8155

Max Act: 2.9022. Min Act: -0.0001

Data Index: 42311 (C4 (Web Text))

Max Activating Token Index: 178

Text #8

Max Range: 4.8155. Min Range: -4.8155

Max Act: 2.6273. Min Act: -0.0001

Data Index: 1214472 (C4 (Web Text))

Max Activating Token Index: 575

Text #9

Max Range: 4.8155. Min Range: -4.8155

Max Act: 2.6325. Min Act: -0.0001

Data Index: 16829 (C4 (Web Text))

Max Activating Token Index: 147

Text #10

Max Range: 4.8155. Min Range: -4.8155

Max Act: 2.8866. Min Act: -0.0001

Data Index: 383551 (C4 (Web Text))

Max Activating Token Index: 702

Text #11

Max Range: 4.8155. Min Range: -4.8155

Max Act: 2.8188. Min Act: -0.0001

Data Index: 855705 (C4 (Web Text))

Max Activating Token Index: 296

Text #12

Max Range: 4.8155. Min Range: -4.8155

Max Act: 2.8171. Min Act: -0.0001

Data Index: 1007699 (C4 (Web Text))

Max Activating Token Index: 11

Text #13

Max Range: 4.8155. Min Range: -4.8155

Max Act: 2.8177. Min Act: -0.0001

Data Index: 1070134 (C4 (Web Text))

Max Activating Token Index: 16

Text #14

Max Range: 4.8155. Min Range: -4.8155

Max Act: 2.6577. Min Act: -0.0001

Data Index: 328574 (C4 (Web Text))

Max Activating Token Index: 237

Text #15

Max Range: 4.8155. Min Range: -4.8155

Max Act: 2.7759. Min Act: -0.0001

Data Index: 693924 (C4 (Web Text))

Max Activating Token Index: 45

Text #16

Max Range: 4.8155. Min Range: -4.8155

Max Act: 2.6280. Min Act: -0.0001

Data Index: 512997 (C4 (Web Text))

Max Activating Token Index: 2

Text #17

Max Range: 4.8155. Min Range: -4.8155

Max Act: 2.7491. Min Act: -0.0001

Data Index: 1057514 (C4 (Web Text))

Max Activating Token Index: 938

Text #18

Max Range: 4.8155. Min Range: -4.8155

Max Act: 2.6155. Min Act: -0.0001

Data Index: 96403 (C4 (Web Text))

Max Activating Token Index: 803

Text #19

Max Range: 4.8155. Min Range: -4.8155

Max Act: 2.7466. Min Act: -0.0001

Data Index: 651399 (C4 (Web Text))

Max Activating Token Index: 823