© 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. 127977
    Prev
    Next
    INDEX
    Explanations

    edges in graphs

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

    The neuron fires on mentions of “edge” or “edges,” i.e. tokens referring to graph edges.

    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
     he
    -3.83
     “
    -2.67
    .”
    -2.63
     is
    -2.52
     (“
    -2.36
    C
    -2.20
    .
    -2.14
    alties
    -2.09
     a
    -2.06
     o
    -2.06
    POSITIVE LOGITS
    geſ
    2.94
     emphatic
    2.73
    谖
    2.70
    ጁ
    2.58
     Each
    2.58
    𖥦
    2.56
    乆
    2.53
    賔
    2.52
    푣
    2.50
     aéreo
    2.48
    Activations Density 0.005%

    No Known Activations