© 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. 117355
    Prev
    Next
    INDEX
    Explanations

    long followed by another descriptor

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

    The neuron activates on occurrences of the word “long,” especially when used adjectivally or in compounds (e.g. “long-term,” “long-chain,” “long-range”).

    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
    ffè
    -0.86
    CAPIT
    -0.77
    řit
    -0.74
    個人的には
    -0.72
    rele
    -0.70
    Excluir
    -0.69
    背影
    -0.68
     Exotic
    -0.68
    сным
    -0.68
    running
    -0.67
    POSITIVE LOGITS
    tua
    0.92
    RANGE
    0.76
    Range
    0.76
    ≔
    0.75
     września
    0.75
    quence
    0.75
    计时
    0.75
    ➞
    0.74
     calloc
    0.74
     Geraadpleegd
    0.74
    Activations Density 0.057%

    No Known Activations