© 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. 107879
    Prev
    Next
    INDEX
    Explanations

    engines and transmissions with speeds and cylinders

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

    The neuron reliably spikes on numeric tokens—especially multi‐digit measurements or specifications (e.g. engine cc, horsepower, model years, gear counts).

    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
    닫
    -0.77
    𝔪
    -0.75
    unif
    -0.75
    phinx
    -0.73
    ikti
    -0.71
    bindingNavigator
    -0.71
    lichte
    -0.71
    wegen
    -0.71
    这些
    -0.70
    annique
    -0.69
    POSITIVE LOGITS
     speed
    1.70
    cylinder
    1.66
     cylinders
    1.66
     cylinder
    1.63
    speed
    1.59
     Cylinder
    1.53
    Cylinder
    1.45
     Speed
    1.36
    cyl
    1.36
     speeds
    1.31
    Activations Density 0.020%

    No Known Activations