© 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 EvalsExportsAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Gemma-2-27B
    5. 22-GEMMASCOPE-RES-131K
    7. 130450
    Prev
    Next
    INDEX
    Explanations

    The neuron strongly activates on section headings or titles—i.e. prominently displayed multi‐word labels (like forum announcements, product names, article headlines).

    oai_token-act-pair · o4-miniTriggered by @jyhe0408

    new

    np_acts-logits-general · gemini-2.5-flash-lite
    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
    </em>
    -1.07
    She
    -1.06
    常用的
    -1.03
     siempre
    -0.99
     aktuell
    -0.99
     gegenwär
    -0.96
     основной
    -0.96
    returns
    -0.95
     zawsze
    -0.90
     verwal
    -0.90
    POSITIVE LOGITS
     new
    1.61
    ใหม่
    1.39
     replacing
    1.31
     will
    1.25
     नए
    1.24
    新的
    1.21
     estrenar
    1.20
     новая
    1.19
     новые
    1.13
     erstat
    1.10
    Activations Density 0.135%

    No Known Activations