© 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. 63138
    Prev
    Next
    INDEX
    Explanations

    names ending in kin or b'erk/b'irk

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

    The neuron fires on any token that contains the letter sequence “kin,” as in proper names like Perkin, Birkbeck, Purkinje, 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
    Ḃ
    -0.81
     bureaus
    -0.69
     детям
    -0.68
    Clash
    -0.67
     competency
    -0.66
    Superficie
    -0.65
     eorum
    -0.65
    mcg
    -0.64
    ビ
    -0.63
    ラスト
    -0.63
    POSITIVE LOGITS
     Vale
    0.83
     만
    0.78
     Ely
    0.74
    nemouth
    0.74
     Stump
    0.73
    FILTER
    0.73
     prema
    0.73
    槽
    0.72
     FILTER
    0.71
     MINT
    0.71
    Activations Density 0.039%

    No Known Activations