© 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. 10-GEMMASCOPE-RES-131K
    7. 7869
    Prev
    Next
    INDEX
    Explanations

    more or longer comparison

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

    The neuron detects quantitative expressions—numbers and number‐related words (e.g. “more,” “longer,” “often,” digit tokens).

    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
    퐿
    -1.37
     sahibi
    -1.35
     Strukt
    -1.34
     eduk
    -1.34
     薬局
    -1.33
    sarung
    -1.32
     geel
    -1.31
     Promotes
    -1.30
     outono
    -1.30
     huel
    -1.29
    POSITIVE LOGITS
     для
    1.70
     when
    1.57
     which
    1.57
     For
    1.56
    The
    1.54
    ).
    1.50
     It
    1.44
    包括
    1.43
     особенно
    1.37
     When
    1.31
    Activations Density 0.023%

    No Known Activations