© 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. 127904
    Prev
    Next
    INDEX
    Explanations

    shooting, killing, stabbing

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

    The neuron strongly activates on past-tense words and phrases describing someone being killed or dying (e.g. “killed,” “shot,” “dead”).

    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
     to
    -1.77
     donde
    -1.67
    ↵↵↵↵↵↵↵
    -1.53
    ↵↵↵
    -1.52
    ↵
    -1.48
     you
    -1.45
     источники
    -1.45
    ↵↵↵↵↵↵↵↵↵↵
    -1.44
    ↵↵↵↵↵↵↵↵↵
    -1.43
    <i>
    -1.43
    POSITIVE LOGITS
    erals
    1.54
    ATIONS
    1.43
    ığımız
    1.41
     néhány
    1.41
    hiva
    1.40
     yksi
    1.37
    Опис
    1.36
    izers
    1.34
    mäßig
    1.32
     parec
    1.32
    Activations Density 0.024%

    No Known Activations