© 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. 57328
    Prev
    Next
    INDEX
    Explanations

    "referred to" or "Although Absolute"

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

    This neuron activates on code‐style tokens—especially XML/HTML tags, attributes, and other programming markers—i.e. it detects segments of document markup or source code rather than ordinary text.

    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
     troubling
    -0.81
     areas
    -0.80
     higher
    -0.79
     identify
    -0.79
     struggling
    -0.78
     striving
    -0.77
    ."
    -0.77
     пробле
    -0.76
     discovering
    -0.75
     venturing
    -0.75
    POSITIVE LOGITS
    astron
    1.04
     különböző
    0.92
    jelly
    0.89
     hebat
    0.89
    xos
    0.85
    棻
    0.84
     berbasis
    0.84
     investimentos
    0.82
     geladen
    0.82
    Approximately
    0.82
    Activations Density 0.000%

    No Known Activations