© 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. 10-GEMMASCOPE-RES-131K
    7. 7553
    Prev
    Next
    INDEX
    Explanations

    symbols preceding specific phrases

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

    This neuron fires most strongly on email‐ or license‐style header tokens and surrounding punctuation—e.g. “To”, “<”, “(” and similar symbols used in metadata 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_10/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
    </em>
    -2.33
     also
    -2.19
    </h2>
    -1.95
     orari
    -1.73
    Also
    -1.70
    There
    -1.69
    時
    -1.64
    會
    -1.63
     with
    -1.60
     歳
    -1.60
    POSITIVE LOGITS
     tratados
    2.00
     
    1.94
     hugely
    1.76
     insanely
    1.76
     acondicionado
    1.75
     comprando
    1.73
     horrifying
    1.71
    犼
    1.71
     brimming
    1.71
     vistos
    1.68
    Activations Density 0.057%

    No Known Activations