© 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. 22-GEMMASCOPE-RES-131K
    7. 128491
    Prev
    Next
    INDEX
    Explanations

    The neuron fires on code tokens involved in numeric/arithmetic operations (e.g. numeric literals, types like “uint,” operators like `*`/`/`, function names like “Mul,” and variable names in expressions).

    oai_token-act-pair · o4-miniTriggered by @jyhe0408

    Spirit studies

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    google/gemma-scope-27b-pt-res/layer_22/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
    Expired
    -0.99
    endcsname
    -0.94
    avez
    -0.91
    一直
    -0.89
    सि
    -0.88
    minY
    -0.88
     you
    -0.84
     mean
    -0.82
     average
    -0.82
    tismo
    -0.82
    POSITIVE LOGITS
     Labyrinth
    1.03
    Confira
    0.95
    してしまう
    0.93
    Both
    0.91
     if
    0.90
    rebbe
    0.90
     এর
    0.89
    📆
    0.87
    entière
    0.87
    imiento
    0.87
    Activations Density 0.005%

    No Known Activations