© 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. 1824
    Prev
    Next
    INDEX
    Explanations

    getter methods

    np_acts-logits-general · gemini-2.5-flash-lite

    The neuron is strongly triggered by tokens used to introduce contact or recipient information (e.g. “mailto:”, “family” in “send flowers to the family,” numeric IDs), i.e. it’s detecting contact‐info markers.

    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
    manifolds
    0.67
    üss
    0.65
     क्रिप्टो
    0.62
    obacterial
    0.62
    ǜ
    0.61
     ReLU
    0.61
    ṑ
    0.60
     这些
    0.59
    äche
    0.59
     చేసి
    0.58
    POSITIVE LOGITS
     toString
    0.58
     Byrne
    0.54
    ='$
    0.54
     varsa
    0.54
     firstName
    0.53
    ખ
    0.52
     unsolicited
    0.50
    ">=</
    0.50
     থাকলে
    0.50
    desPort
    0.49
    Activations Density 0.000%

    No Known Activations