© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    Natural Language
    Autoencoders
    NEW    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Gemma-3-12B
    5. 31-GEMMASCOPE-2-RES-16K
    7. 444
    Prev
    Next
    INDEX
    Explanations

    The neuron is strongly activated by instructional or promotional “how-to” language—tips, advice, and imperative guidance.

    oai_token-act-pair · o4-miniTriggered by @jyhe0408
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    google/gemma-scope-2-12b-pt/resid_post/layer_31_width_16k_l0_medium
    Prompts (Dashboard)
    392,802 prompts, 256 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    No Configuration Found
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    8
    0.49
    F
    0.47
    6
    0.47
    3
    0.45
    5
    0.45
    R
    0.43
    D
    0.43
    4
    0.42
    7
    0.42
    G
    0.42
    POSITIVE LOGITS
     equipments
    0.65
     supaya
    0.64
     walaupun
    0.62
     sufferings
    0.62
     upto
    0.58
     advices
    0.58
     appare
    0.57
     nowadays
    0.57
     sehingga
    0.57
     hence
    0.55
    Activations Density 4.527%

    No Known Activations