© 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. 22-GEMMASCOPE-RES-131K
    7. 34048
    Prev
    Next
    INDEX
    Explanations

    birthday, martyrs'

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

    The neuron strongly responds to capitalized tokens that serve as headings, salutations, or prominent names/terms (e.g. titles like “To,” “Inside,” “Wife,” “Mom,” “Dad,” “Always,” etc.), i.e. words at the start of lines or in card‐style headers.

    oai_token-act-pair · o4-miniTriggered by @jyhe0408
    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
    mogelijk
    -1.05
     posibilidad
    -0.96
    ktı
    -0.96
     пози
    -0.96
     andra
    -0.95
    ίη
    -0.94
    AFP
    -0.93
    南浔
    -0.92
    江苏
    -0.91
    Sz
    -0.91
    POSITIVE LOGITS
    τία
    1.22
     our
    1.15
     насеља
    1.14
     oculta
    1.13
    相处
    1.11
    смот
    1.09
    CARACTER
    1.09
    ственную
    1.07
    辊
    1.02
    REGUL
    1.02
    Activations Density 0.001%

    No Known Activations