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

    Neuronpedia

    Jacobian LensNEW
    Natural Language
    Autoencoders
    NEW    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsinterp-engineNEWAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Gemma-3-12B
    5. 24-GEMMASCOPE-2-RES-16K
    7. 3928
    Prev
    Next
    INDEX
    Explanations

    The neuron strongly activates on the infinitival marker “to” (i.e. “to” plus a verb).

    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_24_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
     that
    0.69
     which
    0.66
     in
    0.64
    я
    0.63
     to
    0.63
     for
    0.63
    in
    0.62
     salah
    0.62
    м
    0.62
     it
    0.62
    POSITIVE LOGITS
    Kết
    0.74
     Kết
    0.65
    妽
    0.61
    Macrophages
    0.58
    Cadastro
    0.57
    벤
    0.56
    Endereco
    0.54
    Coder
    0.53
     Utilisez
    0.53
    Choisissez
    0.53
    Activations Density 0.001%

    No Known Activations