© 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-2-27B
    5. 34-GEMMASCOPE-RES-131K
    7. 78809
    Prev
    Next
    INDEX
    Explanations

    angle: incidence, risk, attack

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

    The neuron is selectively activated by the scientific/technical terms “incident” or “incidence.”

    oai_token-act-pair · o4-miniTriggered by @jyhe0408
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    google/gemma-scope-27b-pt-res/layer_34/width_131k
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    No Configuration Found
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    SDC
    -0.79
    BORN
    -0.73
    ˮ
    -0.72
     العاصمة
    -0.72
     Typhoon
    -0.71
    hirkan
    -0.71
    mán
    -0.71
     Mardi
    -0.70
    ➥
    -0.70
     Jurgen
    -0.69
    POSITIVE LOGITS
     incident
    3.17
     incidence
    3.02
    incident
    2.50
     Incident
    2.47
    incidence
    2.38
     incidents
    2.13
    Incident
    2.11
     incidente
    2.06
     Incidence
    2.05
    Incidence
    1.86
    Activations Density 0.072%

    No Known Activations