© 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 EvalsExportsAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Gemma-2-27B
    5. 34-GEMMASCOPE-RES-131K
    7. 99077
    Prev
    Next
    INDEX
    Explanations

    Joe Avati, SXSW

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

    The neuron activates on isolated numeric tokens or literals (e.g., version numbers, identifiers, numeric constants).

    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
     so
    -0.84
     then
    -0.78
     fairly
    -0.76
    🎇
    -0.75
     multiple
    -0.73
     both
    -0.71
    нци
    -0.71
    %。
    -0.71
     Whatever
    -0.70
     administrativa
    -0.70
    POSITIVE LOGITS
    ⇄
    1.02
     SHOCK
    0.96
     píše
    0.94
     vaš
    0.94
    chauer
    0.93
     trám
    0.92
     WONDER
    0.89
    Ꞌ
    0.88
     Spaniards
    0.88
     vaše
    0.87
    Activations Density 0.000%

    No Known Activations