© 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. 15784
    Prev
    Next
    INDEX
    Explanations

    other subjects or links

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

    The neuron responds strongly to tokens that are purely numeric (digit sequences) or long runs of the “=” character (i.e. section‐separator lines).

    oai_token-act-pair · o4-miniTriggered by @jyhe0408
    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
     it
    -0.97
    ⅔
    -0.93
    REVISTA
    -0.91
    ⅓
    -0.85
    âtre
    -0.85
     mojo
    -0.84
     überhaupt
    -0.81
    kiel
    -0.81
    它的
    -0.80
    vold
    -0.80
    POSITIVE LOGITS
     other
    1.17
     those
    1.00
    ÍTICA
    0.98
    forKey
    0.98
     amelyek
    0.95
     vinaigre
    0.94
     Soccerway
    0.94
    tow
    0.92
    Other
    0.92
    〉
    0.91
    Activations Density 0.005%

    No Known Activations