© 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. 63411
    Prev
    Next
    INDEX
    Explanations

    roles and groups

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

    The neuron fires on words naming particular people or groups (e.g. roles, categories, or demographic labels such as “parents,” “beginners,” “teacher,” “customer,” “member,” etc.).

    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
    達が
    -0.99
    達は
    -0.98
    の方も
    -0.91
    Maven
    -0.84
    たちも
    -0.83
    ± 
    -0.82
     legendary
    -0.81
    🧇
    -0.78
    цах
    -0.77
    Contribu
    -0.76
    POSITIVE LOGITS
    們
    1.80
    们
    1.54
     who
    1.15
     whom
    1.08
    들
    1.02
     souls
    0.97
    みんな
    0.92
     모두
    0.91
    Innen
    0.88
     everyone
    0.88
    Activations Density 0.010%

    No Known Activations