© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    Natural Language
    Autoencoders
    NEW
    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsAPI Community BlogPrivacy & TermsContact
    1. Home
    2. Gemma-2-27B
    3. 34-GEMMASCOPE-RES-131K
    4. 110868
    Prev
    Next
    INDEX
    Explanations

    [![.svg)](

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

    The neuron fires strongly on short alphanumeric strings used as abbreviations (e.g. “svg,” single‐letter county codes like “B,” two‐letter codes like “Bi”) or other terse tokens, i.e. it detects compact abbreviations/file extensions.

    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
     tirs
    -0.86
    タイル
    -0.78
    好吃的
    -0.78
     haj
    -0.77
    giti
    -0.74
     cond
    -0.73
     incompati
    -0.72
    resourceName
    -0.71
    }&
    -0.71
     unmarried
    -0.71
    POSITIVE LOGITS
     zijne
    0.91
     Pří
    0.85
     vertrou
    0.84
     walaupun
    0.82
     high
    0.82
     committees
    0.82
    preuve
    0.80
     tarifleri
    0.80
    nterior
    0.78
    ؟!
    0.78
    Activations Density 0.005%

    No Known Activations