© 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 EvalsExportsAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Gemma-2-27B
    5. 34-GEMMASCOPE-RES-131K
    7. 98922
    Prev
    Next
    INDEX
    Explanations

    found guilty or pleaded guilty

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

    The neuron fires on words reporting formal legal outcomes, especially verbs like “pleaded,” “found,” “convicted,” and similar terms indicating a court judgment or plea.

    oai_token-act-pair · o4-miniTriggered by @jyhe0408
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    google/gemma-scope-27b-pt-res/layer_34/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
     Colombie
    -0.76
    тик
    -0.75
     sweetened
    -0.73
    ~-
    -0.72
     chré
    -0.72
    HandlerContext
    -0.71
    ಗೆ
    -0.71
     fracaso
    -0.69
     الإعلام
    -0.69
    modèle
    -0.69
    POSITIVE LOGITS
     guilty
    6.19
    guilty
    4.63
     Guilty
    4.34
    Guilty
    4.03
     guilt
    3.67
     innocent
    3.13
     Guilt
    2.52
    innocent
    2.44
     Innocent
    2.41
     guil
    2.30
    Activations Density 0.041%

    No Known Activations