© 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. 58787
    Prev
    Next
    INDEX
    Explanations

    describing components of something

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

    The neuron fires on high‐level “list/introduction” cues—words that frame or introduce components (e.g. “agenda,” “includes,” “findings”).

    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
     in
    -2.47
     on
    -2.28
     have
    -2.19
     with
    -2.14
     from
    -2.05
     and
    -1.88
     get
    -1.78
     through
    -1.78
     has
    -1.77
     within
    -1.71
    POSITIVE LOGITS
    ごちそう
    1.83
     devraient
    1.79
    変な
    1.70
    なぜか
    1.70
     Promotes
    1.70
    わたしは
    1.66
     devront
    1.65
    秶
    1.62
     semelh
    1.61
    まさかの
    1.61
    Activations Density 0.099%

    No Known Activations