© 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. 127923
    Prev
    Next
    INDEX
    Explanations

    killing and death

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

    The neuron selectively activates on words describing lethal violence or homicide (e.g., “slaying,” “murdering,” “killing,” “shot,” “stabbing”).

    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
     about
    -1.83
    鹮
    -1.42
    נוּ
    -1.30
     taurus
    -1.24
     represents
    -1.23
     aDecoder
    -1.23
     perciò
    -1.21
     suggests
    -1.20
     dunque
    -1.20
    太多
    -1.18
    POSITIVE LOGITS
    :
    1.62
    _
    1.36
     allé
    1.36
     широко
    1.34
     zeer
    1.32
     onze
    1.29
     heureuse
    1.25
     хорошо
    1.23
     pesso
    1.23
     veya
    1.20
    Activations Density 0.036%

    No Known Activations