© 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. 67398
    Prev
    Next
    INDEX
    Explanations

    Marion or Marian names

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

    The neuron detects occurrences of the proper name “Marion” (or variants like “Marian”) in 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_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
    栳
    -2.92
    
    -2.64
    贗
    -2.61
    ’.
    -2.50
     irmã
    -2.42
    楦
    -2.42
    神話
    -2.34
     juzg
    -2.31
    鬭
    -2.30
     děti
    -2.25
    POSITIVE LOGITS
    {
    3.61
    L
    3.30
    }
    3.20
    b
    3.14
    3.11
    s
    3.08
    7
    2.97
    X
    2.91
    M
    2.88
    The
    2.86
    Activations Density 0.003%
    No Known Activations