© 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. 10-GEMMASCOPE-RES-131K
    7. 8123
    Prev
    Next
    INDEX
    Explanations

    carnage, mayhem, disease, relationship

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

    The neuron fires strongly on the character sequence “op” (as in “photo op,” “opp,” “opportunity,” etc.), i.e. it detects the substring “op” wherever it appears.

    oai_token-act-pair · o4-miniTriggered by @jyhe0408
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    google/gemma-scope-27b-pt-res/layer_10/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
    da
    -2.39
     vertes
    -2.02
    ୌ
    -1.95
    ka
    -1.84
    ca
    -1.74
     antaranya
    -1.67
    re
    -1.60
    na
    -1.59
    db
    -1.56
    ny
    -1.55
    POSITIVE LOGITS
     a
    2.56
     Such
    1.95
     feiert
    1.86
    geox
    1.85
    maline
    1.82
    irsi
    1.77
    ustice
    1.75
    without
    1.75
     Some
    1.74
     exhibición
    1.71
    Activations Density 0.026%

    No Known Activations