© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    Natural Language
    Autoencoders
    NEW    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Gemma-3-12B
    5. 31-GEMMASCOPE-2-RES-16K
    7. 8541
    Prev
    Next
    INDEX
    Explanations

    criminal record

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

    The neuron detects floating-point numeric tokens (decimal numbers, often appearing in parentheses as stats or coordinates).

    oai_token-act-pair · o4-miniTriggered by @jyhe0408
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    google/gemma-scope-2-12b-pt/resid_post/layer_31_width_16k_l0_medium
    Prompts (Dashboard)
    392,802 prompts, 256 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    No Configuration Found
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    निश्ड
    0.54
    ім
    0.53
    mén
    0.53
    溂
    0.53
     مدا
    0.52
     اتصال
    0.50
    浓
    0.50
     દી
    0.50
    моби
    0.49
    淖
    0.49
    POSITIVE LOGITS
     record
    0.78
     undefeated
    0.76
     Improves
    0.69
     records
    0.68
     wins
    0.68
    記錄
    0.68
     Record
    0.67
    Record
    0.66
     improves
    0.66
    record
    0.64
    Activations Density 0.113%

    No Known Activations