© 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. 31647
    Prev
    Next
    INDEX
    Explanations

    gaming software and names

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

    The neuron is selectively activating on number‐like tokens—i.e. any token representing a numeric value (digits).

    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
    父子
    -0.82
    raded
    -0.77
    anoma
    -0.74
    詼
    -0.73
     Futter
    -0.73
     雙
    -0.72
     uncovered
    -0.70
    готовить
    -0.70
     anzunehmen
    -0.70
    (")");
    -0.69
    POSITIVE LOGITS
    MUD
    1.40
     mud
    1.27
     MUD
    1.22
    Mud
    1.17
    mud
    1.15
    mush
    1.00
    Wiz
    0.97
     Mud
    0.96
    wiz
    0.88
    olc
    0.83
    Activations Density 0.038%

    No Known Activations