© 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. 113765
    Prev
    Next
    INDEX
    Explanations

    lying about or to

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

    The neuron activates on words that denote deception or dishonesty (e.g. lie, lying, falsehood, falsification).

    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
     noirs
    -1.71
    メリカ
    -1.71
     nouve
    -1.69
     penchant
    -1.67
    Опис
    -1.67
     作り方
    -1.60
    Người
    -1.59
     teis
    -1.58
     sortes
    -1.58
    ियर
    -1.55
    POSITIVE LOGITS
     about
    2.13
    遢
    1.70
     to
    1.69
     that
    1.66
     resting
    1.63
    but
    1.63
    ";
    1.59
     envía
    1.59
     getNombre
    1.58
     presentará
    1.56
    Activations Density 0.011%

    No Known Activations