© 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. 3035
    Prev
    Next
    INDEX
    Explanations

    ASCII tables and code structures

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

    The neuron activates on horizontal-rule–style separator lines composed of repeated punctuation characters (e.g. long runs of “–” or “=”).

    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
    Weaknesses
    -0.83
    🪚
    -0.82
     trato
    -0.80
    主要的
    -0.78
    それと
    -0.78
    】【
    -0.77
    Sympto
    -0.76
     큰
    -0.74
    kinetic
    -0.74
     menyediakan
    -0.74
    POSITIVE LOGITS
    PyObject
    0.88
    ///
    0.86
    ge
    0.82
    Hunt
    0.81
    Gee
    0.79
    Replace
    0.78
    笼
    0.77
    ㅋ
    0.77
    국
    0.77
    Tree
    0.76
    Activations Density 0.055%

    No Known Activations