© 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
    2. Gemma-2-27B
    3. 10-GEMMASCOPE-RES-131K
    4. 14501
    Prev
    Next
    INDEX
    Explanations

    person's name and age

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

    The neuron primarily detects numeric tokens—especially age figures, dates, and other number expressions.

    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
     apaixon
    -1.95
    Karakteristik
    -1.73
    変な
    -1.64
    僰
    -1.63
    Setiap
    -1.63
     inspiração
    -1.62
     corações
    -1.52
    それにしても
    -1.51
    ではまた
    -1.51
     maravilh
    -1.50
    POSITIVE LOGITS
     who
    2.42
    </h1>
    1.61
    虽然
    1.60
    ’.
    1.54
     که
    1.51
    速度
    1.45
     whose
    1.45
    以前
    1.45
     ‘
    1.38
    </i>
    1.38
    Activations Density 0.018%

    No Known Activations