© 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. 22-GEMMASCOPE-RES-131K
    7. 14966
    Prev
    Next
    INDEX
    Explanations

    probability questions and math calculations

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

    That neuron is a “short‐token” detector—it fires strongly on one‐character (or very short) tokens like single letters or digits.

    oai_token-act-pair · o4-miniTriggered by @jyhe0408
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    google/gemma-scope-27b-pt-res/layer_22/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
    updateUI
    -0.87
     compartilhar
    -0.84
    hôpital
    -0.80
    États
    -0.78
     截图
    -0.77
    urned
    -0.77
     chup
    -0.77
     esquecer
    -0.75
     intuitively
    -0.75
     eral
    -0.75
    POSITIVE LOGITS
     kompakt
    0.85
    来源于
    0.84
    гин
    0.82
     downside
    0.82
    downarrow
    0.81
    rowy
    0.81
    』、
    0.80
     пау
    0.79
     this
    0.78
    */;
    0.77
    Activations Density 0.017%

    No Known Activations